XRCC1 deficiency drives telomeric chromatin leakage, inflammatory signalling and senescence Authors: Anna Piscone, Francesca Gorini, Susanna Ambrosio, Carmen Caiazza, Giannicola Marotta, Anna Noviello, Giovanni Di Fusco, Giovanni Scala, Barbara Majello, Stefano Amente Published: 2026 ABSTRACT Genome integrity is constantly challenged by oxidative DNA damage, which, if left unrepaired, can result in replication stress, chromosomal instability, and tumorigenesis. The base excision repair (BER) pathway is essential for resolving oxidative base lesions such as 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG). Among BER factors, APE1 and XRCC1 play critical roles not only in lesion repair but also in maintaining replication fork stability. Here, we investigate the cellular consequences of APE1 and XRCC1 deficiency in isogenic breast epithelial models (MCF10A and MCF7), under basal and oxidative stress conditions to better dissect the mechanisms through which these factors safeguard genome integrity. We show that XRCC1 deficiency, more than APE1 depletion, leads to an increase of the DNA damage signalling, and the formation of cytoplasmic chromatin fragments (CCFs) containing telomeric DNA. XRCC1-deficient cells also exhibit activation of the cGAS-STING pathway, inflammatory signalling, and senescence phenotypes. Our findings identify XRCC1 as a central suppressor of oxidative stress-induced cytosolic DNA accumulation and immune signalling. These results suggest that targeting XRCC1 may offer a therapeutic avenue to trigger tumour senescence and promote anti-tumour immunity. Introduction The integrity of the genome is persistently threatened by endogenous and exogenous sources of DNA damage. Among these, oxidative stress is a major driver of genome instability, leading to the formation of reactive oxygen species (ROS) that can oxidize DNA bases, producing lesions such as 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodG)—a frequent and mutagenic modification [1,2]. If unrepaired, oxidative DNA damage can stall replication forks and transcriptional machinery, resulting in double-strand breaks (DSBs), chromosomal instability, and potentially oncogenic transformation [3,4]. The base excision repair (BER) pathway is the primary mechanism responsible for removing small base lesions, including 8-oxodG and abasic sites, that do not significantly distort the DNA helix [5]. The repair of 8-oxodG is initiated by OGG1 DNA glycosylases that recognize and excise damaged bases, followed by APE1 (apurinic/apyrimidinic endonuclease 1), which cleaves the phosphodiester backbone at the resulting abasic sites. Repair is completed by DNA polymerase β, XRCC1 (X-ray repair cross-complementing protein 1)—a scaffold protein that coordinates polymerase and ligase recruitment—and DNA ligase IIIα [[6], [7], [8]]. Beyond their canonical roles, APE1 and XRCC1 are critical in the response to replication stress. XRCC1-deficient cells accumulate single-strand breaks (SSBs), which can be converted into DSBs during S-phase [9]. XRCC1 is also required for replication fork progression and restart, and its absence leads to replication fork collapse and increased chromosomal fragmentation [10]. Similarly, impaired APE1 activity leads to accumulation of toxic BER intermediates, fork stalling, checkpoint activation, and genomic instability [11]. Telomeres, being rich in guanine residues, are highly susceptible to oxidative damage [12]. Unrepaired lesions at telomeres hinder replication and can promote end-to-end fusions, chromosomal aberrations, or the generation of telomere-containing cytoplasmic chromatin fragments (CCFs) [13]. Both XRCC1 and APE1 play protective roles at telomeres. XRCC1-mediated BER activity is required to repair oxidative base damage and prevent telomere dysfunction [14], while APE1 has been shown to regulate telomere integrity both enzymatically and through redox control of telomere-associated factors [15]. Oxidative stress and elevated ROS levels are potent inducers of cellular senescence, a stable cell cycle arrest accompanied by a pro-inflammatory secretory phenotype, chromatin remodelling, and metabolic changes [[16], [17], [18]]. Senescence acts as a tumour-suppressive mechanism by preventing the proliferation of cells with damaged genomes, but in tumour cells it can have dual effects: while limiting growth of damaged cells, senescent tumour cells can secrete factors (SASP, senescence-associated secretory phenotype) that promote for instance inflammation and tumour progression [19,20]. Notably, although 8-oxodG and BER proteins are known to trigger senescence [13,21], the precise mechanisms linking BER dysfunction to senescence induction remain to be better understood. Recent studies have linked the presence of 8-oxodG at gene regulatory regions to the transcription of pro-inflammatory genes and revealed epigenetic roles for 8-oxodG and BER proteins [2,[22], [23], [24], [25], [26], [27], [28], [29]]. OGG1 and APE1 regulate gene expression also by modulating the redox status of transcription factors such as NF-κB, p53, and AP-1, thereby integrating DNA repair with transcriptional responses [30]. Moreover, BER machinery operates within chromatin and requires chromatin remodelling for efficient access to nucleosome DNA [31,32]. However, unresolved 8-oxodG repair intermediates in active chromatin regions, or in proximity of replication origins contained in long genes [22], may provoke replication fork stalling and generate DSBs, linking epigenetic regulation to genome instability [33]. Despite these advances, the role of APE1 and XRCC1 in safeguarding genome integrity remains to be clarified. In this study, we investigated the mechanisms by which loss of APE1 or XRCC1 impacts genome stability in non-tumorigenic (MCF10A) and tumorigenic (MCF7) breast epithelial cell models. We found that XRCC1 deficiency, more than APE1 loss, promotes accumulation of telomere-positive cytoplasmic chromatin fragments which in turn drive inflammatory activation, and senescence. Methods Cell culture and cell lines generation MCF10A cells (non-transformed mammary epithelial cell line) were maintained as previously reported [22,25,26]. Briefly, MCF10A cells were cultured in DMEM/F-12 medium (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 5% horse serum, 20 ng/mL epidermal growth factor (EGF), 100 ng/mL cholera toxin, 10 μg/mL insulin, and 500 ng/mL hydrocortisone. MCF7 cells (human breast adenocarcinoma cell line) were cultured in Dulbecco's Modified Eagle Medium (DMEM; Gibco) lacking sodium pyruvate, supplemented with 10% fetal bovine serum (FBS; Gibco). APE1-and XRCC1-deficient MCF10A and MCF7 cell lines were generated using the CRISPR/Cas9 system. Lenti-X 293T cell line (Takara, 632180) was used to produce lentiviral particles using standard procedures. Briefly, lentiviruses were produced by transfection of Lenti-X 293T cells with the “all-in-one” pLentiCRISPR-V2 (Addgene, #52961) plasmid, which expresses S. pyogenes Cas9 and guide RNAs (gRNA), along with packaging plasmids pMD2G-VSV-G (Addgene, #12259) and psPAX2 (Addgene, #12260), using Lipofectamine 2000 transfection reagent (Invitrogen, 11668019). Sixteen hours post transfection, medium was replaced with DMEM supplemented with 20% FBS. After 48 h, virus-containing supernatants were collected, filtered through a 0.45 μm nylon membrane to eliminate cell debris and concentrated with polyethylene glycol (PEG). Specifically, PEG-8000 (final concentration 10% w/v) and NaCl (final concentration 0.3 M) were added to the filtered supernatant, mixed thoroughly, and incubated at 4°C overnight to allow viral particles to precipitate overnight. Then, the mixture was centrifuged at 1600 × g for 60 min at 4°C to pellet the viral particles. The resulting pellet was gently resuspended in 1/100 of the original supernatant volume using cold DMEM/F12 supplemented with 1% Penicillin-Streptomycin and stored at −80°C in aliquots until use. For viral transduction, recipient cells were seeded at a density of 4 × 105 cells in a 60 mm dish, and infected with lentiviral particles diluted 1:50 in the presence of polybrene (Millipore, TR-1003-G) at a final concentration of 10 μg/mL. Media was changed 24 h post-infection, and cells were selected using puromycin (Sigma-Aldrich, 540411) at a final concentration of 1 μg/mL. The expression of targeted proteins was evaluated by western blotting. gRNA sequences with low off-target scores were selected using the CHOPCHOP online tool (https://chopchop.cbu.uib.no/). One gRNA was chosen for each gene; a non-targeting gRNA was used to generate control cells (CTRL) under the same experimental conditions. Selected gRNAs are listed below.gAPEX1: GATCAGAAAACCTCACCCAG;gXRCC1: GAATGATGGCTCAGCTTTCG;gCTRL: GTAGCGAACGTGTCCGGCGT. gAPEX1: GATCAGAAAACCTCACCCAG; gAPEX1: GATCAGAAAACCTCACCCAG; gXRCC1: GAATGATGGCTCAGCTTTCG; gXRCC1: GAATGATGGCTCAGCTTTCG; gCTRL: GTAGCGAACGTGTCCGGCGT. gCTRL: GTAGCGAACGTGTCCGGCGT. All cell lines were maintained under standard conditions at 37°C in a humidified incubator with 5% CO2. BER-deficient cells were maintained as pooled populations to minimize clonal selection bias. For the UV treatment, exponentially growing cells were irradiated with 254-nm UV light at 40 J/m2. Cell viability assay by MTT Cell viability was assessed using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay Kit (Invitrogen V13154). MCF10A cells (CTRL, APE1, and XRCC1-deficient cells) were seeded in 96-well plates at a density of 103 cells/well in 200 μL of complete medium and allowed to adhere overnight. Cells were then treated with the indicated concentrations of potassium bromate (KBrO3) and subsequently allowed to recover in fresh culture medium for the indicated time. After treatment and recovery, 10 μL of an MTT stock solution (12 mM in phosphate-buffer saline, PBS) were added to each well in 100 μL of complete medium and cells were incubated for 4 h at 37°C under standard culture conditions. Following incubation, 85 μL of incubation medium were removed, and the resulting formazan crystals were solubilized by adding 50 μL of solubilization buffer (DMSO) to each well for 10–15 min at 37°C in the dark. Absorbance was measured at 540 nm using a microplate reader (Biotek Sinergy H1, AHSI). Cell viability was quantified by normalizing the absorbance values obtained for each time point and experimental condition to those of the corresponding untreated control at time 0. Each condition was analysed in technical triplicates and in at least three independent biological replicates. Data were analysed using GraphPad Prism v8.4.2 software and are reported as mean ± SEM. 8-oxodG immunofluorescence microscopy Detection of 8-oxodG was performed using a mouse monoclonal anti-8-oxodG antibody (clone 15A3; Trevigen, cat# 4354-MC-050). Specifically, MCF10A cells were seeded at a density of 1.5x105 cells per well in 12-well plates, each containing a 12 × 12 mm coverslip, and allowed to adhere overnight. Cells were then washed with PBS and fixed with an ice-cold methanol-acetone solution for 20 min at −20°C, then allowed to air dry. Next, DNA denaturation was performed by incubating the samples in 0.05 N HCl for 5 min on ice, followed by PBS washes. RNA was subsequently digested by incubating the cells with RNase A (100 μg/ml in 150 mM NaCl, 15 mM sodium citrate) for 1 h at 37°C. The cells were then washed sequentially in PBS-diluted ethanol solutions at progressively higher concentrations (35%, 50% and 75% respectively), followed by in situ DNA denaturation with 0.15 N NaOH in 70% ethanol for 4 min. After that, cells were briefly rinsed with PBS and stained with Hoechst 33342 (0.2 μg/ml in PBS; Immunochemistry Technologies) for 10 min. Following this step, cells were treated with 4% methanol-free formaldehyde solution (Thermo Fisher Scientific, 28908) in 70% ethanol and gradually rehydrated through sequential incubations in 50% and 35% ethanol containing 4% formaldehyde, followed by PBS washes. Next, cells were incubated with 5 μg/ml proteinase K, diluted in a 20 mM Tris, 1 mM EDTA (pH 7.5) solution, for 10 min at 37°C. To prevent non-specific antibody binding, cells were incubated with 5% normal goat serum solution in PBS for 1 h at room temperature (RT). The cells were then incubated overnight at 4°C with the primary anti-8-oxodG antibody in PBS containing 1% BSA and 0.01% Tween-20, in a humidified chamber. Following washing with PBS containing 0.05% Tween-20 (0.05% TPBS), the cells were incubated with Alexa Fluor 488-conjugated goat anti-mouse secondary antibody in PBS containing 1% BSA for 1 h at room temperature in the dark. Finally, after additional washes in 0.05% TPBS, samples were then mounted onto glass microscope slides using ProLong™ Glass Antifade Mountant (Thermo Fisher Scientific) and imaged using a confocal microscope. Immunofluorescence microscopy Cells were seeded at defined densities based on the specific experimental requirements on coverslips and treated as indicated. Following treatment and/or recovery, cells were washed twice with PBS and fixed with 4% methanol-free formaldehyde solution (Thermo Fisher Scientific, 28908) in PBS. MCF10A cells were permeabilized with 0.5% Triton X-100 in PBS at RT (0.1% Triton X-100 was used for MCF7 cells) and blocked with 3% bovine serum albumin (BSA) and 0.1% Tween-20 in PBS at RT under gentle agitation. Cells were incubated with primary antibodies for 2 h at RT or overnight at 4°C, depending on the target antigen. Subsequently, cells were washed with PBS before incubating with the appropriate Alexa Fluor-conjugated secondary antibodies (Life Technologies, USA) at RT in the dark. Nuclear counterstaining was performed using DAPI (Invitrogen™, #62248) at a final concentration of 2 μg/mL at RT. Coverslips were then mounted onto glass microscope slides using ProLong™ Glass Antifade Mountant (Thermo Fisher Scientific) and imaged using a fluorescence or confocal microscope. Antibodies used were rabbit monoclonal anti-γH2AX (clone JBW301, #ab81299, Abcam), rabbit polyclonal anti-53BP1 (Elabscience, E-AB-61102), rabbit monoclonal anti-cGAS (clone D1D3G, #15102S, Cell Signaling Technology, USA). BrdU assay Cells were seeded onto sterile glass coverslips at a density of 3 × 105 cells each. After seeding, cells were allowed to adhere and then pulse-labeled with 10 μM BrdU (Sigma-Aldrich, B5002) for 90 min. Potassium bromate (KBrO3, 10 mM) was added 30 min before the end of the pulse. Following treatment, the growth medium was removed, and cells were washed three times with PBS. Cells were then fixed in 4% methanol-free formaldehyde solution (Thermo Fisher Scientific, 28908) in PBS for 15 min. After fixation, cells were quenched with 0.5 M glycine (Sigma-Aldrich, 219517) in PBS for 15 min and permeabilized in 0.5% Triton X-100 (Sigma-Aldrich, T8787) in PBS at RT. Subsequently, the DNA denaturation was carried out using 2 N HCl (Carlo Erba 403876) for 15 min; cells were then blocked with 2% BSA and 0.1% Tween-20 in PBS under gentle agitation at RT. Then, cells were incubated with anti-BrdU primary antibody (Thermo Fisher Scientific, MA1-81890) overnight at 4°C. Next day, cells were washed with PBS and incubated with anti-γH2AX Ser139 (Abcam, ab81299) for 2 h at RT. Finally, cells were washed with PBS and incubated in the dark at room temperature with Alexa Fluor 594-conjugated anti-mouse IgG secondary antibody (Thermo Fisher Scientific, A-21203) for BrdU detection and Alexa Fluor 488-conjugated anti-rabbit IgG secondary antibody (Thermo Fisher Scientific, A-21206) for γH2AX (Ser139) detection. Nuclear counterstaining was performed using 4′,6- diamidino-2-phenylindole (DAPI) (Thermo Fisher Scientific, 62248) at a final concentration of 2 μg/mL. Coverslips were then mounted onto glass microscope slides using ProLongTM Glass Antifade Mountant (Thermo Fisher Scientific, P36980) and imaged using confocal microscopy. Immunofluorescence and telo-fluorescence in situ hybridization (IF-tF) MCF10A cells were seeded at defined densities based on the specific experimental requirements in 12-well plates, each containing a 12 × 12 mm coverslip. Following treatment and/or recovery, cells were washed twice with PBS and fixed with 4% methanol-free formaldehyde solution (Thermo Fisher Scientific, 28908) in PBS at RT. Subsequently, cells were permeabilized with 0.5% Triton X-100 in PBS at RT and then incubated with a blocking buffer solution containing 5% normal goat serum, 1% BSA, and 0.1% Triton X-100 in PBS. Primary antibodies were diluted in blocking buffer and incubated overnight at 4°C. The following antibodies were used: rabbit anti-53BP1 (Elabscience, E-AB-61102) and mouse anti-γH2AX (Santa Cruz Biotechnology, sc-517348). The following day, cells were washed with PBS and incubated for 1 h at room temperature in the dark with fluorophore-conjugated secondary antibodies (Alexa Fluor 647 goat anti-rabbit IgG and Alexa Fluor 594 goat anti-mouse IgG) diluted in blocking buffer. After incubation, the cells were washed with a 0.2% Triton solution in PBS. To preserve fluorescence, a second fixation step with 4% formaldehyde for 12 min at RT was performed. Then, cells were rinsed with 1% BSA and washed with PBS. For FISH analysis, cells underwent sequential dehydration in 70%, 90%, and 100% ethanol for 5 min each. During dehydration, the telomeric PNA probe was diluted 1:100 (PNA Bio, F1004, TelC-Alexa 488, 3xCCCTAA) in hybridization buffer (70% formamide, 10 mM Tris HCl pH 7.5, 1x Maleic Acid buffer, 1x MgCl2 buffer). The probe solution was heated at 80°C for 5 min, then cooled on ice for 2 min. Following ethanol dehydration, coverslips were air-dried and inverted onto microscope slides containing the probe solution. Hybridization was carried out on a hot plate at 80°C for 10 min and then at room temperature for 2 h in a humidified chamber protected from light. After hybridization, coverslips were washed twice for 15 min in a 70% formamide, 10 mM Tris-HCl (pH 7.5) solution on a rocking platform, followed by 5 min washes in PBS-Triton and rinses in PBS. Then, nuclei were counterstained with DAPI (2 μg/mL in PBS; Merck) for 15 min at RT. Coverslips were mounted on microscope slides using ProLong™ Glass Antifade Mountant (Thermo Fisher Scientific) and visualized using a confocal fluorescence microscope. Senescence assay Cells were seeded in 6-well plates at a density appropriate to reach the desired confluency after 72 h of culture under standard conditions. At the experimental endpoint, cells were gently washed twice with PBS and fixed using the Fixative Solution provided in the Senescence β-Galactosidase Staining Kit (Cell Signalling Technology). After fixation, cells were washed again with PBS and incubated with the freshly prepared β-galactosidase staining solution, following the manufacturer's instructions. The staining solution was applied directly to the cells, and the plates were placed in an incubator at 37°C (without CO2) O/N or 24 h for MCF7 or MCF10A, respectively. Following incubation, cells were examined under a bright-field microscope. Senescent cells were identified by the presence of blue precipitate localized in the cytoplasm and quantified as indicated below. Image acquisition and analysis For the analysis of γH2AX and 53BP1 foci, fluorescence images were acquired using a confocal microscope (Zeiss LSM 700) with appropriate filter settings to detect the respective fluorescence signals. Foci quantification was performed using the Foci Analyzer plugin in Fiji (ImageJ), which enabled automated detection and measurement of foci within the cells. The parameters for identifying γH2AX and 53BP1 foci—specifically, the minimum foci size (in pixels/voxels) and the intensity threshold bias—were empirically determined using the MCF10A-AsiSI-ER cell line as a reference. This model is known to form γH2AX and 53BP1 foci in response to DNA DSBs upon induction with 400 nM 4-hydroxytamoxifen (OHT) for 4 h [34]. Supplementary Fig. 1 shows representative images of γH2AX and 53BP1 foci formation in MCF10A-AsiS1-ER cells, validating the robustness and specificity of the used parameters. For each experimental condition, at least 100 cells were counted, and each experiment was performed in triplicate to ensure reproducibility. Statistical analyses of foci data were conducted using the Kruskal–Wallis test followed by Dunn's multiple comparisons test (GraphPad Prism v8.4.2) to compare differences between experimental groups. For the visualization of CCFs and analysis of γH2AX or cGAS-positive CCFs, images were acquired using a fluorescence or confocal microscope. Quantitative analyses, including statistical comparisons, were performed using GraphPad Prism v8.4.2 (GraphPad Software, USA). Statistical analysis was performed using one-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test), to assess differences between experimental groups. Data are reported as mean ± standard deviation (SD) For IF-tF acquisition and analysis, to prevent signal intensity dependency on cell aggregation and/or uneven spread onto slides, images were taken throughout the slide avoiding imaging aggregates. Image acquisition parameters (gain, digital offset, digital gain and pinhole) for both DAPI, Alexa fluor 647, 594 and 488 were kept consistent across experimental conditions in each separate experiment. The percentage of area showing the signal intensity was calculated using Fiji (ImageJ). Each image was first converted to 8-bit grayscale using Image → Type → 8-bit. A threshold was applied using Image → Adjust → Threshold. Once the appropriate threshold was set, it was applied to binarize the image. To measure the proportion of the image area above the threshold, the option ‘Area Fraction’ was selected under Analyze → Set Measurements. Then, the measurement was obtained using Analyze → Measure, which provided the percentage of the total image area above the intensity threshold. This value directly represents the %Area reported in the results table. For BrdU and γH2AX co-staining analysis, raw data from immunofluorescence images was extracted with ImageJ/Fiji (v.2.16.0/1.54p), using the custom macro described as follows. First, the channels from each image were split and slices were stacked using Max Intensity. Then, applying the Li dark threshold, the DAPI staining was used to obtain the nuclear mask and to create a ROI Manager entry for each nucleus. Only particles with a minimum size of 20 μm2 were considered as nuclei. Next, the brightness and contrast of the Alexa fluor 488 and Alexa fluor 594 channels were adjusted to remove the background signal. Then, the Default dark threshold was applied and further fine-tuned to match the background-corrected image. The final threshold values were fixed across the images from the same experimental set. To detect the foci of γH2AX a Watershed binary adjustment was applied before counting particles satisfying a minimum threshold size of 0.01 μm2. For BrdU quantification, we measured the fraction of nuclear area positive for the signal (area of signal/area of nucleus x 100). The threshold for BrdU positivity was defined based on the negative control population, and cells were classified as BrdU-positive when the area fraction exceeded a fixed cut-off (≥1.4%), which was consistently applied across all conditions. Protein extraction and western blotting For total protein extraction, cells were scraped and resuspended in Triton buffer solution (1% Triton X-100, 150 mM NaCl, 50 mM Tris-HCl pH 7.5, 5 mM EDTA, and protease and phosphatase inhibitor cocktails; Thermo Fisher). Samples were incubated for 30 min on ice, and the supernatant was collected by centrifugation at 18,400 x g for 11 min at 4°C to remove debris. Protein concentrations were determined using the Bradford Protein Assay Kit (Bio-Rad) following the manufacturer's instructions. Equal amounts of total protein extract were mixed with 3x Blue Loading Buffer and stored at −80°C until use. Prior to electrophoresis, samples were denatured by heating at 95°C for 5 min. Proteins were separated by SDS-PAGE and subsequently transferred onto a nitrocellulose membrane at 4°C. Membranes were then washed with either PBST (PBS supplemented with 0.1% Tween-20) or TBST (TBS supplemented with 0.1% Tween-20) and blocked at room temperature with 5% nonfat dry milk or 3% BSA, prepared in the corresponding washing buffer (PBST or TBST). Primary antibodies were diluted in the same blocking solution and incubated overnight at 4°C under gentle agitation. After washing, membranes were incubated at room temperature with HRP-conjugated secondary antibodies diluted in the blocking solution. Protein bands were visualized using an ECL substrate according to the manufacturer's instructions, and signal acquisition was performed using the ChemiDoc™ MP imaging system (Bio-Rad) or by developing the membranes with X-ray film using a film developer. Band intensities were quantified using ImageJ software (NIH) or Image Lab software (BioRad), and relative protein expression was normalized to the loading control (GAPDH or actinin). The following primary antibodies were used: rabbit polyclonal anti-APE1 (FNab00480, FineTest), rabbit anti-XRCC1 (NBP1-87154, Novus), mouse anti-GAPDH (Santa Cruz Biotechnology, sc-365062), mouse monoclonal anti-α-Actinin (sc-17829, Santa Cruz Biotechnology), rabbit monoclonal anti-STING (clone D2P2F, #13647, Cell Signaling Technology, USA), and rabbit monoclonal anti-phospho-STING (Ser366) (clone E9A9K, #50907, Cell Signaling Technology, USA), rabbit polyclonal anti-p21 (c-19, sc397, Santa Cruz Biotechnology). Full scan of the western blots is reported in Supplementary Figs. 11 and 12. RNA-sequencing and analysis Total RNA was extracted from MCF10 CTRL, APE1-and XRCC1-deficient cells using the RNeasy Mini Kit (QIAGEN, Hilden, Germany), according to the manufacturer's protocol. RNA extraction was performed in biological triplicates. The concentration and purity of the RNA were initially assessed using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, USA). RNA-seq libraries and sequencing were performed using Illumina® Stranded mRNA Prep and sequenced with Novaseq 6000. Sequences were processed using the ARMOR pipeline version 1.5.7 [35]. In particular, trimming and adapter removal was performed using Trim Galore (https://github.com/FelixKrueger/TrimGalore) version 0.6.6 with 20 as Phred score cut-off, gene level read quantification was performed using Salmon version 1.4.0 over GRCh37 (hg19) reference transcriptome annotation with --seqBias --gcBias --fldMean 250 --fldSD 25 parameters, differential expression analysis was performed using edgeR version 3.34.1 [36] by filtering out low expression genes using the filterByExpr edgeR function and correcting p-values using FDR. Gene Set Enrichment Analysis (GSEA) was performed using the GSEA software package (Broad Institute, v4.3.2) with gene sets from the Molecular Signatures Database (MSigDB), including Hallmark, Gene Ontology (GO) Biological Process, and KEGG pathways. The analysis was conducted on differentially expressed genes to identify significantly enriched biological pathways, molecular functions, and cellular components. OxiDIP-seq and analysis Genomic DNA fragments enriched for 8-oxodG from untreated and UV-treated MCF10A cells were obtained as previously described [22,25,26,37]. OxiDIP-seq data were analysed following established protocols [22,25,26,37]. Briefly, raw sequencing reads were subjected to quality control using the NGS-QC Toolkit [38], then aligned to the reference genome using BWA [39] with default parameters. Subsequent filtering and file format conversions were performed using SAMtools [40] and Bedtools [41]. Data visualization was conducted via the UCSC Genome Browser. For signal analysis, uniquely mapped reads were normalized to genomic input using the bamCompare tool from the DeepTools suite [42] applying the SES (Signal Extraction Scaling) method (32) as a normalization factor. This approach accounts for GC-content–related sequencing bias and differences in DNA input between samples and controls. Metagene profiles and heatmaps were generated using the computeMatrix and plotHeatmap functions within DeepTools. Signal profile plots were created in R, based on matrices output by computeMatrix. Statistical analysis The number of technical and biological replicates are reported in all Figures legends and methods. The applied statistical test is indicated in Figures legends. All statistical tests were made using R (v.4.2.2) and GraphPad Prism (v.8.4.2, GraphPad Software, San Diego, CA, USA). All the statistical analyses have been conducted using 0.05 as statistical significance threshold. Results APE1 and XRCC1 deficiencies cause an increase of 8-oxodG levels Based on the canonical function of APE1 and XRCC1 in the 8-oxodG repair process, APE1-deficient cells are expected to accumulate unrepaired AP sites, whereas XRCC1-deficient cells are likely to accumulate unresolved SSBs. Hence, we used CRISPR-Cas9 technology to disrupt APEX1 or XRCC1 genes, generating APE1-or XRCC1-deficient cells, respectively. As shown in Fig. 1A and Supplementary Fig. 2, APEX1 knockout resulted in a highly efficient and homogeneous cell population with complete protein loss, while XRCC1 knockout yielded a slightly more heterogeneous population comprising a substantial fraction of fully knocked-out cells, alongside some heterozygous as well as very few potentially wild-type cells (Fig. 1A and Supplementary Fig. 2). This discrepancy may reflect differences in the essentiality of these genes: XRCC1 has been reported to be critical for cell viability, with complete XRCC1 loss leading to early embryonic lethality in mice and reduced survival in somatic cells [8,43,44]. Therefore, cells retaining partial XRCC1 expression—either due to heterozygosity or in-frame indels preserving partial function—may be preferentially retained over time, explaining the persistence of a mixed population even under constitutive Cas9 expression.Fig. 1BER deficiency promotes accumulation of nuclear 8-oxodG.(A) Western blot analysis of APE1 and XRCC1 expression in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells. GAPDH was used as a loading control. (B) MTT assay measuring cell viability in CTRL, APE1-deficient, and XRCC1-deficient MCF10A cells at the indicated time points. Each condition was analysed in technical triplicates and in three independent biological replicates. Data are presented as mean ± SEM. Statistical differences between experimental groups were evaluated using two-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test). Only statistically significant differences are reported on the graph. (C) Representative images (left panel) of CTRL, APE1-deficient, and XRCC1-deficient MCF10A cell nuclei stained with anti-8-oxodG (green) and Hoechst (blue); scale bars = 5 μm. Violin plot (right panel) shows quantification of 8-oxodG levels as mean fluorescence intensity/nucleus. n = number of nuclei analysed from two independent experiments. Solid lines represent the median; dashed lines indicate the first and third quartiles. Statistical differences were evaluated using the Kruskal–Wallis test followed by Dunn's multiple comparisons test in GraphPad Prism v8.4.2 (GraphPad Software, San Diego, CA, USA). Only statistically significant differences are reported in the graph. (D) MTT assay of CTRL, APE1-deficient, and XRCC1-deficient MCF10A cells following recovery at the indicated time points after induced oxidative stress with 30′ of KBrO3 (0, 10, 25, and 50 mM) treatment. Each condition was analysed in technical triplicates and in three independent biological replicates. Data are presented as mean ± SEM. Statistical differences were evaluated using two-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test). Only statistically significant differences are reported on the graph. BER deficiency promotes accumulation of nuclear 8-oxodG.(A) Western blot analysis of APE1 and XRCC1 expression in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells. GAPDH was used as a loading control. (B) MTT assay measuring cell viability in CTRL, APE1-deficient, and XRCC1-deficient MCF10A cells at the indicated time points. Each condition was analysed in technical triplicates and in three independent biological replicates. Data are presented as mean ± SEM. Statistical differences between experimental groups were evaluated using two-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test). Only statistically significant differences are reported on the graph. (C) Representative images (left panel) of CTRL, APE1-deficient, and XRCC1-deficient MCF10A cell nuclei stained with anti-8-oxodG (green) and Hoechst (blue); scale bars = 5 μm. Violin plot (right panel) shows quantification of 8-oxodG levels as mean fluorescence intensity/nucleus. n = number of nuclei analysed from two independent experiments. Solid lines represent the median; dashed lines indicate the first and third quartiles. Statistical differences were evaluated using the Kruskal–Wallis test followed by Dunn's multiple comparisons test in GraphPad Prism v8.4.2 (GraphPad Software, San Diego, CA, USA). Only statistically significant differences are reported in the graph. (D) MTT assay of CTRL, APE1-deficient, and XRCC1-deficient MCF10A cells following recovery at the indicated time points after induced oxidative stress with 30′ of KBrO3 (0, 10, 25, and 50 mM) treatment. Each condition was analysed in technical triplicates and in three independent biological replicates. Data are presented as mean ± SEM. Statistical differences were evaluated using two-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test). Only statistically significant differences are reported on the graph. Next, to assess whether APE1 knockout and XRCC1 knockdown affect cell proliferation in our BER-deficient cells, we performed an MTT assay at 0, 24, 48, 72 and 96 h. As shown in Fig. 1B, differences in cell viability were observed between BER-deficient (APE1 and XRCC1) and Control (CTRL) cells, with a statistically significant 24% cell growth reduction at 96 h for APE1-deficient cells (30.7 vs 40.2 abs, APE1 and CTRL, respectively) and 38% for XRCC1-deficient cells (24.8 vs 40.2 abs, XRCC1 and CTRL, respectively). These data indicate that deficiency of APE1 or XRCC1 may impair cell growth under basal conditions. Furthermore, to evaluate the impact of BER deficiency on the repair of oxidative DNA damage — and considering that both APE1 and XRCC1 have been reported to modulate OGG1 activity [[45], [46], [47]] — we quantified the levels of 8-oxodG by immunofluorescence. As shown in the violin plot in Fig. 1C, both APE1-and XRCC1-deficient cells exhibited a marked increase in nuclear 8-oxodG levels compared to Control cells, with 1850, 3840, and 4570 fluorescence intensity median values for CTRL, APE1-, and XRCC1-deficient cells, respectively (Fig. 1C). These results suggest that the earlier step of the 8-oxodG repair process is impaired without functional APE1 or XRCC1, consistent with the knowledge that OGG1 catalytic activity is enhanced via interactions with APE1 [48] and XRCC1 [45]. These findings prompted us to evaluate cell proliferation of our BER-deficient cells under conditions of induced oxidative stress. To this end, we treated BER-deficient cells with 0, 10, 25, and 50 mM of potassium bromate (KBrO3) - a well-established oxidative agent that primarily induces 8-oxodG lesions in DNA [49,50] - for 30 min. Samples were collected at the recovery time points ranging from 0 to 96 h as indicated in Fig. 1D. Cell viability assay, assessed by MTT (Fig. 1D) in Control, APE1-, and XRCC1-deficient cells, showed a dose- and time-dependent reduction in proliferation. In particular, at 96 h, treatment with 10 mM KBrO3 resulted in a statistically significant 30% cell growth reduction for APE1-deficient cells (24.7 vs 35.0 abs, APE1 and CTRL, respectively) and 56% for XRCC1-deficient cells (15.2 vs 35.0 abs, XRCC1 and CTRL, respectively) compared to treated Control cells (Fig. 1D), whereas higher concentrations (25 and 50 mM) of KBrO3 induced strong cytotoxic effects in all cell types. This supports that our BER-deficient cells are less capable of coping with oxidative DNA damage, leading to slight impaired proliferation under stress conditions. Overall, these findings suggest that XRCC1- or APE1-deficiency impairs cell proliferation under basal conditions in association with increased level of endogenous 8-oxodG. However, the BER-deficient cells exhibit increased sensitivity to oxidative stress, confirming the critical role of XRCC1 and APE1 in protecting against oxidative DNA damage. APE1 and XRCC1 deficiencies cause the formation of DSBs Uncompleted repair of 8-oxodG can lead to the formation of SSBs, which may subsequently convert into DSBs [[2], [3], [4]]. Hence, we next investigated whether in APE1-and XRCC1-deficient cells also accumulate DNA strand breaks. To test this, we performed immunofluorescence staining using antibodies against histone H2AX phosphorylated at Ser-139 (γH2AX) to detect γH2AX foci in Control, APE1-, and XRCC1-deficient cells cultured under basal growth conditions or exposed to oxidative stress with 10 mM KBrO3 for 30’. Under normal growth conditions, Control cells exhibited a mean of 2.6 γH2AX foci per nucleus, whereas APE1-and XRCC1-deficient cells displayed increased mean values of 3.3 and 3.8 foci per nucleus, respectively (Fig. 2A). Among the BER-deficient models, only XRCC1-deficient cells showed a statistically significant increase in γH2AX foci compared to Control, indicating an increase of the DNA damage in the absence of XRCC1.Fig. 2APE1 and XRCC1 deficiencies activate the DNA damage response (DDR).(A) Representative images (left panel) of Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells under normal growth conditions (−) or after induced oxidative stress with 30′ of KBrO3 (+), stained with anti-γH2AX (green) and DAPI (blue); scale bars = 5 μm. Plot (right panel) shows quantification of γH2AX foci. n = number of nuclei analysed from three independent experiments. Statistical differences were evaluated using the Kruskal–Wallis test followed by Dunn's multiple comparisons test. Only statistically significant differences are reported in the graph. (B) Representative images (left panel) of CTRL, APE1-deficient, and XRCC1-deficient MCF10A cells under normal growth conditions (−) or after induced oxidative stress with 30′ of KBrO3 (+), stained with anti-53BP1 (red) and DAPI (blue); scale bars = 5 μm. Plot (right panel) shows quantification of 53BP1 foci. n = number of nuclei analysed from three independent experiments. Statistical differences were evaluated using the Kruskal–Wallis test followed by Dunn's multiple comparisons test. Only statistically significant differences are reported on the graph. APE1 and XRCC1 deficiencies activate the DNA damage response (DDR).(A) Representative images (left panel) of Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells under normal growth conditions (−) or after induced oxidative stress with 30′ of KBrO3 (+), stained with anti-γH2AX (green) and DAPI (blue); scale bars = 5 μm. Plot (right panel) shows quantification of γH2AX foci. n = number of nuclei analysed from three independent experiments. Statistical differences were evaluated using the Kruskal–Wallis test followed by Dunn's multiple comparisons test. Only statistically significant differences are reported in the graph. (B) Representative images (left panel) of CTRL, APE1-deficient, and XRCC1-deficient MCF10A cells under normal growth conditions (−) or after induced oxidative stress with 30′ of KBrO3 (+), stained with anti-53BP1 (red) and DAPI (blue); scale bars = 5 μm. Plot (right panel) shows quantification of 53BP1 foci. n = number of nuclei analysed from three independent experiments. Statistical differences were evaluated using the Kruskal–Wallis test followed by Dunn's multiple comparisons test. Only statistically significant differences are reported on the graph. Upon exposure to oxidative stress (10 mM KBrO3 for 30 min), γH2AX foci increased in all cell types (mean values: 3.7, 4.1, and 5.3 foci per nucleus for Control, APE1-, and XRCC1-deficient cells, respectively;Fig. 2A). Notably, XRCC1-deficient cells exhibited a significantly greater increase in γH2AX foci number relative to both APE1-deficient and Control cells. To further validate these findings, we analysed the formation of 53BP1 foci. Interestingly, we observed under basal growth conditions a significant increase in 53BP1 foci in both APE1-and XRCC1-deficient cells (per nucleus foci mean value = 3.5 and 7.6 for APE1-and XRCC1-deficient cells respectively; Fig. 2B) compared to Controls (per nucleus foci mean value = 1.6; Fig. 2B). Upon 10 mM of 30’ KBrO3 exposure, 53BP1 foci increased in all cell lines (foci mean value = 7.6, 7.5 and 9.9 for Control, APE1-and XRCC1-deficient cells, respectively; Fig. 2B) compared to respective untreated cells, with the most pronounced accumulation again in XRCC1-deficient cells even compared to APE1-deficient cells (Fig. 2B). The slight discrepancy between γH2AX and 53BP1 staining in APE1-deficient cells may reflect differences in the sensitivity or dynamics of these two proteins. The observed accumulation of γH2AX and 53BP1 foci in BER-deficient cells suggests that unrepaired BER intermediates may accumulate and be converted into DSBs during replication. Thus, to assess whether DNA damage preferentially arises in replicating cells, we performed BrdU pulse labelling followed by BrdU/γH2AX co-immunostaining. This analysis revealed that BER-deficient cells, and particularly XRCC1-deficient cells, compared to Control cells displayed more γH2AX foci preferentially in BrdU-positive cells (Fig. 3 and Supplementary Fig. 3), either under untreated conditions or following KBrO3 treatment, indicating that the observed DNA damage is largely replication-associated.Fig. 3APE1 and XRCC1 deficiencies promote replication-associated DNA damage response activation. (A) Quantification of the number of γH2AX foci within positive (+) or negative (−) BrdU nuclei in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells grown under basal condition and after induced oxidative stress (10 mM KBrO3 for 30′). Data are shown as mean ± SEM from two independent experiments; n = total number of cells analysed. P-values were calculated using one-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test), to assess differences between experimental groups in GraphPad Prism. (B) Representative images of BrdU-positive nuclei with γH2AX foci in cells as described in (A). APE1 and XRCC1 deficiencies promote replication-associated DNA damage response activation. (A) Quantification of the number of γH2AX foci within positive (+) or negative (−) BrdU nuclei in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells grown under basal condition and after induced oxidative stress (10 mM KBrO3 for 30′). Data are shown as mean ± SEM from two independent experiments; n = total number of cells analysed. P-values were calculated using one-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test), to assess differences between experimental groups in GraphPad Prism. (B) Representative images of BrdU-positive nuclei with γH2AX foci in cells as described in (A). Together, these results suggest that deficiencies in core BER factors, particularly XRCC1, impair the resolution of oxidative base damage and lead to persistent DNA damage signalling and replication-associated stress under both basal and oxidative stress conditions, with DSBs representing a possible downstream consequence. XRCC1 deficiency triggers telomere-positive CCFs formation Recent studies have shown that OGG1 and MUTYH activities promote telomere dysfunction leading to the accumulation of senescence-associated extranuclear DNA in the form of CCFs [13]. CCFs are cytoplasmic structures that are γH2AX-positive but 53BP1-negative by immunofluorescence. Although they resemble micronuclei (MN), CCFs differ in their origin: while MN typically arise during mitosis, CCFs form outside of mitosis, likely through the blebbing of damaged chromatin into the cytoplasm [51]. To investigate whether impaired APE1 or XRCC1 activity promotes CCF formation through chromatin leakage caused by unresolved DNA damage, we examined CCFs accumulation in cells deficient for each of these BER factors by analysing γH2AX and 53BP1 signals using an immunofluorescence-based (IF-tF) assay. As shown in Fig. 4A and quantified in Fig. 4B, under basal growth conditions, CCFs were detected in 2.2% of Control cells, 5.7% of APE1-deficient cells, and 11.8% of XRCC1-deficient cells. While APE1-deficient cells exhibited a slight (2.6-fold up) increase in CCFs frequency compared to Control, XRCC1-deficient cells showed a marked (5.3-fold up) increase, suggesting that deficiency of XRCC1 strongly promotes chromatin leakage under basal conditions. Similar results were obtained 48 h after oxidative stress, induced by 30 min treatment with 10 mM KBrO3, with the frequency of CCF-positive cells increased to 8.9% in Control, 10.8% in APE1-deficient, and 21.5% in XRCC1-deficient cells (Fig. 4B). The 48-h recovery period was chosen to allow accumulation of cytoplasmic chromatin fragments, which arise as a delayed consequence of unresolved DNA damage. These results indicate that XRCC1 knockdown, more than APE1 knockout, exacerbates the accumulation of DNA damage-associated CCFs both under basal and stress conditions, consistent with a more severe impairment in DNA repair and chromatin integrity.Fig. 4XRCC1 deficiency induces DDR-positive cytoplasmic chromatin fragments (CCFs). (A) Schematic representation and representative images of cytoplasmic DNA species stained by IF-tF with DAPI (blue), anti-γH2AX (gray), and anti-53BP1 (red); scale bars = 10 μm. (B) Quantification of the percentage of cells containing cytoplasmic CCFs in the indicated cell lines under basal (solid bars) and after 48 h of recovery time from induced oxidative stress (striped bars). Data in the panel are shown as mean ± SEM from four independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction), to assess differences between experimental groups in R. Only statistically significant differences are reported on the graph. (C) Quantification of the percentage of the indicated cells displaying CCFs-positive for γH2AX or 53BP1 signal under basal growth condition (solid bars) and after induced oxidative stress with 30′ of KBrO3 (striped bars). Data are shown as mean ± SEM from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction), to assess differences between experimental groups in R. Only statistically significant differences are reported on the graph. XRCC1 deficiency induces DDR-positive cytoplasmic chromatin fragments (CCFs). (A) Schematic representation and representative images of cytoplasmic DNA species stained by IF-tF with DAPI (blue), anti-γH2AX (gray), and anti-53BP1 (red); scale bars = 10 μm. (B) Quantification of the percentage of cells containing cytoplasmic CCFs in the indicated cell lines under basal (solid bars) and after 48 h of recovery time from induced oxidative stress (striped bars). Data in the panel are shown as mean ± SEM from four independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction), to assess differences between experimental groups in R. Only statistically significant differences are reported on the graph. (C) Quantification of the percentage of the indicated cells displaying CCFs-positive for γH2AX or 53BP1 signal under basal growth condition (solid bars) and after induced oxidative stress with 30′ of KBrO3 (striped bars). Data are shown as mean ± SEM from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction), to assess differences between experimental groups in R. Only statistically significant differences are reported on the graph. Moreover, with normal growth conditions, co-staining with anti-γH2AX and 53BP1 antibodies revealed that a high fraction of CCFs were positive for γH2AX staining specifically in XRCC1-deficient cells compared to Control cells (mean value = 1.0, 2.0 and 4.8 percent for Control, APE1-and XRCC1-deficient cells respectively; Fig. 4A and C). In contrast, these CCFs remained largely negative for 53BP1 across all cell types, including Control, APE1-, and XRCC1-deficient cells (Fig. 4A and C), consistent with the canonical phenotype of CCFs. After 48 h from oxidative stress (10 mM KBrO3 for 30’), a similar pattern was observed: XRCC1-deficient cells displayed an increase in γH2AX-positive CCFs compared to Control cells (mean value = 4.4, 6.2 and 9.8 percent for Control, APE1-and XRCC1-deficient cells respectively; Fig. 4C), while 53BP1 remained largely low from CCFs in all conditions. These findings suggest that XRCC1-deficiency promotes the formation of damaged chromatin fragments and their subsequent accumulation in cytoplasm. Next, considering that telomeres are particularly vulnerable to oxidative damage [12]— especially the formation of 8-oxodG, which compromises their stability — and that the repair of 8-oxodG at telomeres has been implicated in the generation of telomere-containing CCFs [13], we first examined whether oxidative lesions may accumulate at telomeres in parental MCF10A cells. To this end, we took advantage of previously generated OxiDIP-seq datasets obtained from MCF10A cells under basal conditions and following UV-induced oxidative stress, a treatment known to promote both genome-wide 8-oxodG formation [22] and CCFs generation (Supplementary Fig. 4A). Metagene analysis over a 100 kb window surrounding telomeric regions revealed a clear enrichment of 8-oxodG at telomeres (Fig. 5A), which was further confirmed by genome browser inspection (Supplementary Fig. 5). Intriguingly, while OxiDIP-seq analyses of UV-treated MCF10A cells revealed increased 8-oxodG levels at specific regulatory genomic regions compared with untreated cells [22,25,26], the same datasets did not show a further enrichment of 8-oxodG at telomeric regions (Fig. 5A, compare 8-oxodG signal before and after UV treatment). This observation is consistent with the possibility that induced oxidative stress, promoting telomere dysfunction and fragmentation, may lead to a telomere shortening [52,53] and a relocation of a significative fraction of telomeric DNA into CCFs (Supplementary Fig. 4B) and, consequently, underrepresentation of oxidized nuclear telomeric sequences captured by OxiDIP-seq.Fig. 5XRCC1 deficiency induces telomere-containing CCFs. (A) Plots showing metagene analysis of mean 8-oxodG signal, normalized over the input DNA, across a 100 kb telomeric window in untreated and UV-treated MCF10A cells. The 95% confidence interval (mean ± 2 standard error) of the mean is indicated by the light-blue shaded areas. A representative chromosome image with telomeric regions is shown below. Red arrows mark 8-oxodG enrichment. (B) A representative image of a nucleus stained by IF-tF with DAPI (blue), telomeric probe (green) and anti-53BP1 (red); scale bars = 5 μm (left panel). Quantification of the average number of 53BP1 foci colocalizing with telomeres (TIFs) in nucleus of Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells grown under basal conditions (solid bars) and after induced oxidative stress (10 mM KBrO3 for 30′) (striped bars), stained by IF-tF (right panel). Data are shown as mean ± SD from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction. (C) Quantification of cells displaying CCFs-positive for telomeric DNA in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells under basal conditions (solid bars) and after 48 h of recovery time from induced oxidative stress (10 mM KBrO3 for 30′) (striped bars), stained by IF-tF with DAPI (blue), telomeric probe (TTAGGG, green) anti-γH2AX (gray), and anti-53BP1 (red). Data are shown as mean ± SD from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction). XRCC1 deficiency induces telomere-containing CCFs. (A) Plots showing metagene analysis of mean 8-oxodG signal, normalized over the input DNA, across a 100 kb telomeric window in untreated and UV-treated MCF10A cells. The 95% confidence interval (mean ± 2 standard error) of the mean is indicated by the light-blue shaded areas. A representative chromosome image with telomeric regions is shown below. Red arrows mark 8-oxodG enrichment. (B) A representative image of a nucleus stained by IF-tF with DAPI (blue), telomeric probe (green) and anti-53BP1 (red); scale bars = 5 μm (left panel). Quantification of the average number of 53BP1 foci colocalizing with telomeres (TIFs) in nucleus of Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells grown under basal conditions (solid bars) and after induced oxidative stress (10 mM KBrO3 for 30′) (striped bars), stained by IF-tF (right panel). Data are shown as mean ± SD from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction. (C) Quantification of cells displaying CCFs-positive for telomeric DNA in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells under basal conditions (solid bars) and after 48 h of recovery time from induced oxidative stress (10 mM KBrO3 for 30′) (striped bars), stained by IF-tF with DAPI (blue), telomeric probe (TTAGGG, green) anti-γH2AX (gray), and anti-53BP1 (red). Data are shown as mean ± SD from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using Fisher's exact test, followed by multiple comparison post hoc tests (with Bonferroni correction). Our data indicate that BER deficiency leads to the accumulation of oxidative DNA damage and increased DNA damage signalling, which are associated with the formation of telomere-containing CCFs, thereby potentially promoting telomere instability. To test this hypothesis, we performed a Telomere dysfunction Induced Foci (TIF) analysis [54] by using IF-tF assay. Hence, we assessed colocalization of 53BP1 foci with telomeres in nucleus of Control and BER-deficient cells, either under basal growth conditions or after induced oxidative stress with 30’ of 10 mM of KBrO3. After inducing oxidative stress, more than 44% of APE1-deficient cells and 55% of XRCC1-deficient cells displayed at least one TIF, as opposed to 14% of Control cells (Fig. 5B). In contrast, under basal conditions, no significant differences were observed. Moreover, under induced oxidative stress, APE1-deficient cells exhibited up to 7 TIFs per nucleus, whereas XRCC1-deficient cells showed up to 8 TIFs per nucleus (Supplementary Fig. 6) compared to 2 TIFs per nucleus in Control cells. These results confirm that telomeres are particularly susceptible to oxidative DNA damage [12] in BER-deficient cells following induced oxidative stress. Based on this observation, we next investigated whether BER deficiency leads to accumulation of telomeric DNA in CCFs – a clear marker of telomere instability – by analysing the telomeric FISH signal of IF-tF assay performed under basal conditions and 48 h after induced oxidative stress with 30’ of 10 mM KBrO3. Notably, APE1-and XRCC1-deficient cells exhibited an increase in telomere-positive CCFs in both untreated and treated cells (mean values: 1.6%, 3.8%, and 5.7% in Control, APE1-, and XRCC1-deficient cells under normal growth; 3.7%, 6.0%, and 14.4% following induced oxidative stress; Fig. 5C). Moreover, we found particularly under oxidative stress that 56% and 67% of CCFs result positive for telomeres in APE1-and XRCC1-deficient cells, respectively, compared with 42% observed in Control cells (compare Figs. 4B and 5C). These results indicate that the deficiency of XRCC1, and to a lesser extent APE1, leads to the cytoplasmic leakage of telomeric DNA, consistent with unresolved telomeric damage and defective BER activity at telomeres. XRCC1 deficiency promotes cGAS-STING activation and proinflammatory response Cytoplasmic DNA, particularly in the form of CCFs, is known to activate the innate immune sensor cyclic GMP-AMP synthase (cGAS) and the STING pathway [51]. Hence, we first assessed the presence of cGAS-positive CCFs in BER-deficient cells. Immunofluorescence staining with anti-cGAS antibodies was performed under basal conditions and 72 h (corresponding to 24 h following the time of CCF detection, as cGAS activation is a downstream event of CCF formation) after oxidative stress induced with a 30’ of 10 mM KBrO3. As shown in Fig. 6A, we detected an increased number of cGAS-positive CCFs in APE1-and particularly in XRCC1-deficient cells under basal conditions (mean value = 0, 0.35, 1.05 percent for Control, APE1-and XRCC1-deficient cells, respectively), with a more pronounced accumulation upon oxidative stress (mean value = 2.0, 3.9 and 4.5 percent for Control, APE1-and XRCC1-deficient cells, respectively).Fig. 6XRCC1 deficiency triggers activation of the cGAS-STING pathway and an Inflammatory response. (A) Schematic representation and representative images of Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells 72 h after induced oxidative stress (10 mM KBrO3 for 30′), stained with cGAS (green) and DAPI (blue); scale bars = 5 μm. Arrows highlight extranuclear CCFs that can be either cGAS-positive (green arrow) or cGAS-negative (white arrow). Plot showing quantification of the indicated cells displaying CCFs-positive for cGAS. Data are shown as mean ± SD from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using one-way analysis of variance (ANOVA) on arcsine square root-transformed data, followed by multiple comparison post hoc tests (Tukey's test). (B) Western blot analysis of STING, p-STING protein levels in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells, treated with KBrO3, and harvested after the recovery time from induced oxidative stress indicated in the schematic representation. Actinin was used as a loading control. Plot (right panel) reports, as mean ± SEM from three independent experiments, the quantification of the p-STING protein level normalized over Actinin. (C) Gene set enrichment analysis showing Normalized Enrichment Scores (NES) for the top 5 positively and negatively enriched genes sets (FDR <0.05) in XRCC1- or APE1-deficient versus Control cells. A positive NES indicates enrichment in the BER-deficient cells, a negative NES indicates enrichment in the Control cells. XRCC1 deficiency triggers activation of the cGAS-STING pathway and an Inflammatory response. (A) Schematic representation and representative images of Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells 72 h after induced oxidative stress (10 mM KBrO3 for 30′), stained with cGAS (green) and DAPI (blue); scale bars = 5 μm. Arrows highlight extranuclear CCFs that can be either cGAS-positive (green arrow) or cGAS-negative (white arrow). Plot showing quantification of the indicated cells displaying CCFs-positive for cGAS. Data are shown as mean ± SD from three independent experiments; n = total number of cells analysed. Statistical differences were evaluated using one-way analysis of variance (ANOVA) on arcsine square root-transformed data, followed by multiple comparison post hoc tests (Tukey's test). (B) Western blot analysis of STING, p-STING protein levels in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF10A cells, treated with KBrO3, and harvested after the recovery time from induced oxidative stress indicated in the schematic representation. Actinin was used as a loading control. Plot (right panel) reports, as mean ± SEM from three independent experiments, the quantification of the p-STING protein level normalized over Actinin. (C) Gene set enrichment analysis showing Normalized Enrichment Scores (NES) for the top 5 positively and negatively enriched genes sets (FDR <0.05) in XRCC1- or APE1-deficient versus Control cells. A positive NES indicates enrichment in the BER-deficient cells, a negative NES indicates enrichment in the Control cells. Then, to further validate activation of the STING pathway, we conducted Western blot analysis to assess levels of STING and its activated phosphorylated form (p-STING). As shown in Fig. 6B, both APE1-and XRCC1-deficient cells exhibited elevated levels of p-STING relative to Control cells under basal conditions, with a further increase detected 96 h (corresponding to 24 h following the time of cGAS-positive CCF detection, as phosphorylation of STING is a downstream event of cGAS activation) after oxidative stress induced by 30 min treatment with 10 mM KBrO3. Collectively, these findings support the hypothesis that unresolved DNA damage in BER-deficient cells, particularly in the absence of XRCC1, promotes activation of the cGAS-STING signalling axis. Since activation of the STING pathway drives the transcriptional induction of pro-inflammatory and immune-related genes, we next explored the broader transcriptional consequences of BER deficiency by performing RNA sequencing (RNA-seq), followed by differential expression and GSEA analyses, on Control, APE1-and XRCC1-deficient cells. XRCC1-deficient cells exhibited significant upregulation of gene sets associated with oxidative phosphorylation, TNF-α signalling via NF-κB, p53 pathway, DNA repair among others, accompanied by marked downregulation of pathways involved in the G2/M checkpoint, E2F targets, and Mitotic spindle (Fig. 6C and Supplementary Fig. 7). In contrast, APE1-deficient cells showed transcriptional activation of the p53 pathway, mTORC1 signalling, and UV response gene sets, along with suppression of IL-6/JAK/STAT3 signalling, TNF-α signalling via NF-κB, and additional inflammatory pathways. These data suggest that, although both BER deficiencies trigger DNA damage responses, XRCC1 deficiency is more strongly associated with inflammatory signalling than APE1 loss. XRCC1 deficiency fosters cellular senescence phenotypes Given that BER deficiencies lead to DSBs and the formation of telomere-positive CCFs and considering that CCFs are recognized hallmarks of cellular senescence [51], we next investigated whether deficiency of APE1 or XRCC1 triggers senescence in MCF7 breast cancer cells. MCF7 cells are commonly used as the tumour cell counterpart to the non-tumorigenic MCF10A cells employed in our earlier experiments. To this end, we generated APE1-and XRCC1-deficient MCF7 cell lines, as confirmed by Western blot analysis using anti-APE1 and anti-XRCC1 antibodies (Supplementary Fig. 8A). In BER-deficient MCF7 cells, we recapitulated the increased accumulation of γH2AX foci under basal conditions through immunofluorescence staining, consistent with elevated DNA damage levels (Supplementary Fig. 8B). We next assessed the formation of cGAS-positive CCFs by immunofluorescence under both basal and oxidative stress conditions. XRCC1-deficient cells displayed a significant increase in cGAS-positive CCFs, whereas APE1-deficient cells showed only a modest and not significant increase (Supplementary Fig. 9A). Consistently, Western blot analysis revealed elevated levels of phosphorylated STING (p-STING), particularly in XRCC1-deficient cells (Supplementary Fig. 9B). Finally, to directly assess cellular senescence, we performed senescence-associated β-galactosidase (SA-β-gal) staining in both MCF10A and MCF7 Control, APE1-, and XRCC1-deficient cells under basal conditions and 72 h (corresponding to the detection time of telomere-positive CCFs) after oxidative stress induced with a 30’ treatment with 10 mM KBrO3. Strikingly, tumorigenic MCF7 cells exhibited clear signs of senescence phenotype, particularly in the XRCC1-deficient background. Specifically, XRCC1-deficient MCF7 cells, compared to the Control cells, showed a marked increase in SA-β-gal positivity, approximately of 2-fold or 4-fold up under oxidative stress or basal conditions respectively. In contrast, APE1-deficient MCF7 cells displayed senescence levels comparable to those of Control cells (Fig. 7A and B). This phenotype was further supported by an increase of p21 protein levels in XRCC1-deficient cells compared to the Control cells (Fig. 7C). Notably, p21 levels were also increased in APE1-deficient cells upon oxidative stress, despite the absence of a corresponding increase in SA-β-gal staining. This apparent discrepancy may reflect differences in the regulation and kinetics of senescence-associated pathways, as p21 induction can occur as part of a DNA damage response that does not necessarily culminate in a full senescent phenotype. In contrast, XRCC1 deficiency appears to promote a more sustained and robust senescence phenotype.Fig. 7XRCC1 deficiency induces senescent phenotype in MCF7 cells. (A) Representative images of β-galactosidase positive cells obtained from Control (CTRL), APE1-deficient, and XRCC1-deficient MCF7 cells under normal conditions (−) and after 72 h of recovery time from induced oxidative stress with 30′ of 10 mM KBrO3 (+) as indicated in schematic representation. Red arrows mark positive cells. (B) Plot reports the percent β-galactosidase positive area of the experiment represented in A. Data are mean ± SD from three independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test), in GraphPad Prism. (C) Western blot analysis of p21 protein levels in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF7 cells, treated with KBrO3, and harvested after the recovery time from induced oxidative stress indicated in the schematic representation. Actinin was used as a loading control. Plot (right panel) reports, as mean ± SEM from three independent experiments, the quantification of the p21 protein level normalized over Actinin. XRCC1 deficiency induces senescent phenotype in MCF7 cells. (A) Representative images of β-galactosidase positive cells obtained from Control (CTRL), APE1-deficient, and XRCC1-deficient MCF7 cells under normal conditions (−) and after 72 h of recovery time from induced oxidative stress with 30′ of 10 mM KBrO3 (+) as indicated in schematic representation. Red arrows mark positive cells. (B) Plot reports the percent β-galactosidase positive area of the experiment represented in A. Data are mean ± SD from three independent experiments. Statistical analysis was performed using one-way analysis of variance (ANOVA), followed by multiple comparison post hoc tests (Tukey's test), in GraphPad Prism. (C) Western blot analysis of p21 protein levels in Control (CTRL), APE1-deficient, and XRCC1-deficient MCF7 cells, treated with KBrO3, and harvested after the recovery time from induced oxidative stress indicated in the schematic representation. Actinin was used as a loading control. Plot (right panel) reports, as mean ± SEM from three independent experiments, the quantification of the p21 protein level normalized over Actinin. Similar results were obtained for BER-deficient MCF10A cells (Supplementary Fig. 10), although the senescent phenotype appeared markedly less pronounced than that observed in MCF7 cells. Our findings suggest that oxidative DNA damage, together with cGAS-STING pathway engagement, drives activation of senescence-associated phenotypes in XRCC1-deficient cells, thereby highlighting its function in preventing both inflammatory response and senescence. Discussion Disruption of base excision repair (BER) genes compromises genomic integrity and elicits a robust cellular response to oxidative stress [5]. In this study, we corroborate the knowledge that APE1 and XRCC1 deficiencies impair the processing of oxidative DNA lesions [[6], [7], [8]] and provide evidence that reduction of XRCC1 protein levels exerts a broader impact on telomere stability, cytosolic DNA accumulation, inflammatory signalling, and senescence. Our data confirm that APE1 and XRCC1 are critical for limiting the accumulation of oxidative base damage. BER-deficient cells show increased 8-oxodG levels under basal conditions, consistent with impaired repair of oxidized guanine, a major product of reactive oxygen species (ROS). Notably, this damage modestly impairs proliferation under basal conditions, suggesting the presence of compensatory mechanisms that transiently preserve cellular fitness despite defective BER. Upon oxidative stress, however, BER-deficient cells exhibit increased vulnerability, with a marked reduction in viability and proliferation, corroborating the importance of APE1 and XRCC1 in safeguarding cells under genotoxic conditions. Among the two, XRCC1 deficiency has a more pronounced impact, leading to an increase of DNA damage signalling, as evidenced by the accumulation in γH2AX and 53BP1 foci following oxidative stress. This is consistent with XRCC1's key role as a scaffold protein coordinating the repair of SSBs [55] that can otherwise collapse into DSBs if unrepaired. A striking feature of XRCC1-deficient cells is the accumulation of telomere-positive CCFs. These structures, containing γH2AX foci but devoid of 53BP1, likely reflect damaged and unrepaired chromatin leaking into the cytoplasm [51]. Given the known sensitivity of telomeres to oxidative damage [12], our finding that XRCC1 deficiency, and to a lesser extent APE1, leads to telomeric CCFs supports the notion of a specialized role for XRCC1 in telomere maintenance. Importantly, CCF accumulation occurred under both basal and stress conditions, suggesting a chronic state of genome instability in XRCC1-deficient cells. At the nuclear level, our data show that damaged telomeres are marked by colocalization of 53BP1 foci. However, a subset of telomeres appears to be released as CCFs lacking 53BP1, suggesting defective resolution of DSBs repair in the absence of XRCC1. Further studies will be necessary to elucidate the molecular mechanisms underlying this defect. Consistent with the accumulation of extranuclear DNA, we observed activation of the cGAS-STING innate immune pathway, particularly in XRCC1-deficient cells. This activation led to increased phosphorylation of STING and upregulation of inflammatory transcriptional programs, including TNF-α signalling, oxidative phosphorylation, and DNA damage response pathways. Interestingly, APE1-deficient cells activated only a subset of these responses and failed to show substantial engagement of inflammatory signalling. This difference may reflect either (i) activation of distinct downstream cGAS–STING signalling pathways or additional inflammatory mechanisms in XRCC1-deficient cells, or (ii) the activation of different inflammatory programs by cGAS–STING signalling in APE1-deficient cells. Notably, XRCC1 deficiency induces marked senescence-associated phenotypes in tumour cells (MCF7) compared to non-tumorigenic epithelial cells (MCF10A). Although senescence has been well described in cancer cells [19,20], the stronger induction of senescence phenotypes in XRCC1-deficient MCF7 cells, compared to XRCC1-deficient MCF10A cells, may reflect intrinsic differences between normal and cancer cells in terms of stress tolerance, DNA repair proficiency, or telomere dynamics. This observation highlights a potential tumour-specific vulnerability to XRCC1 loss [56] and may open new therapeutic opportunities as current therapeutic strategies targeting BER defects mainly rely on PARP inhibitors to selectively target tumour cells [[57], [58], [59]]. The mechanistic basis for this response may involve breast cancer-specific telomere regulation, higher replication stress, or differential reliance on XRCC1-mediated repair, and warrants further mechanistic investigation. Our results may have several important implications. First, they confirm that APE1 and XRCC1 deficiencies sensitize cells to oxidative stress by causing the accumulation of unresolved DNA damage and repair intermediates, with a particular impact at telomeric regions. These lesions may be converted into more deleterious DNA structures during replication, thereby contributing to persistent DNA damage signalling and to immune activation. Notably, XRCC1 appears to play a role in limiting these effects, as its loss is associated with enhanced cytoplasmic DNA accumulation, cGAS–STING activation, and senescence. Second, the more pronounced senescence phenotypes observed in tumour cells may raise the possibility of exploiting XRCC1 loss or BER inhibition to trigger anti-tumour responses while minimizing effects on normal tissues. This strategy may be particularly effective in combination with immunotherapies, as the activation of cGAS-STING can enhance tumour immunogenicity and immune cell recruitment. Importantly, XRCC1 deficiency, which leads to the accumulation of SSBs — but not APE1 deficiency, which results in apurinic site accumulation — may induce a pronounced telomere instability phenotype. This supports the notion that telomeres do not merely act as passive targets of oxidative stress but also as dynamic sensors of deficiencies in the DNA repair mechanism. In this context, telomeres may function as a biological rheostat by gradually converting the burden of unresolved oxidative damage into proportional cellular outcomes, such as DNA damage signalling, cytosolic DNA release, inflammatory activation, and ultimately, senescence. These observations align with the model proposed by the Opresko group, which suggests that it is not the damage sensing itself but the persistence of unrepaired SSB intermediates that drives telomere instability, ultimately activating inflammatory pathways and/or cellular senescence [13,21,60,61]. In summary, our findings identify XRCC1 as a guardian of telomeric stability, especially under oxidative stress. Its loss may promote accumulation of DSBs and telomeric CCFs, chronic cGAS-STING activation, and tumour-specific senescence. From a therapeutic point of view, this suggests that targeting BER —especially in XRCC1-proficient tumours— could sensitize malignant cells to oxidative stress, induce senescence, and potentially boost anti-tumour immunity. These insights provide a strong rationale for combining BER inhibition with pro-oxidant therapies or immune checkpoint blockade to enhance cancer treatment strategies. Funding The research leading to these results has received funding from AIRC under IG 2025 - ID. 31964 project – P.I. Amente Stefano. This research also was supported by “Programma per il Finanziamento della Ricerca di Ateneo Linea A” (FRA) 2022 (University of Naples Federico II) to S.A (Amente) and National Center for Gene Therapy and Drugs based on RNA Technology—PNRR, MUR, Italy to S.A (Amente); National Center for Gene Therapy and Drugs based on RNA Technology—PNRR, MUR, Italy and AIRC IG 23066 to B.M. CRediT authorship contribution statement Anna Piscone: Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing. Francesca Gorini: Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing. Susanna Ambrosio: Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing. Carmen Caiazza: Data curation, Investigation, Writing – review & editing. Giannicola Marotta: Data curation, Formal analysis, Writing – review & editing. Anna Noviello: Data curation, Formal analysis. Giovanni Di Fusco: Data curation, Formal analysis, Writing – review & editing. Giovanni Scala: Data curation, Formal analysis, Writing – review & editing. Barbara Majello: Funding acquisition, Writing – review & editing. Stefano Amente: Conceptualization, Data curation, Funding acquisition, Investigation, Supervision, 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:Amente Stefano reports article publishing charges and equipment, drugs, or supplies were provided by AIRC and PNRR, MUR, Italy. Majello Barbara reports equipment, drugs, or supplies was provided by AIRC and PNRR, MUR, Italy. 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.