Type I interferon signaling defines a novel disease signature in xeroderma pigmentosum C human keratinocytes

Abstract

Xeroderma Pigmentosum C (XPC) is a DNA damage recognition protein central to the global genome nucleotide excision repair (GG-NER) pathway, where it acts as a primary sensor of UV-induced DNA lesions. Loss-of-function mutations in the XPC gene lead to a photosensitive phenotype, with marked accumulation of unrepaired DNA damage and a dramatically elevated risk (10,000-fold) of skin cancer. However, understanding the molecular signaling mechanisms associated with XP-C has been hindered by the lack of reproducible disease models. Here, we overcome this challenge using our genetically engineered human XPC knockout (KO) keratinocytes, the predominant cell type affected by UV radiation. To uncover upstream signaling changes associated with the absence of XPC expression, we quantified protein tyrosine kinase (PTK) activity one-hour post-UVB exposure. XPC KO keratinocytes showed significant dysregulation of PTK activity on ~ 100 phosphosites compared to controls. Complementary mass spectrometry (MS)-based quantitative proteomic analysis performed 24 h post-UVB exposure identified a downstream signature comprising 791 differentially expressed proteins in XPC KO cells irradiated compared to non-irradiated counterparts. An integrative bioinformatic assessment of the kinase activity and proteomic data revealed a significant perturbation in type I interferon signaling via the JAK/STAT pathway in XPC-deficient keratinocytes, which is further exacerbated by UVB exposure. These findings were validated by western blot analysis, establishing a novel disease-associated molecular signature. Given the central role played by JAK/STAT signaling in inflammatory processes, our results implicate this pathway as a key mediator of XP-C’s hypersensitivity, thereby highlighting its potential to alleviate the disease pathology.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-37662-0.

Introduction

The skin is no longer viewed as a passive physical barrier but rather as an active neuro-immuno-endocrine organ that continuously senses environmental cues and translates them into local and systemic homeostatic responses. Recent advances in neuro-immuno-endocrinology of the skin have revealed that keratinocytes, melanocytes, immune cells, and cutaneous nerve endings form an integrated signaling network capable of producing and responding to neurotransmitters, cytokines, hormones, and stress mediators. Through this network, environmental stressors such as ultraviolet (UV) radiation initiate complex signaling cascades that extend beyond DNA damage responses, engaging inflammatory, immune, and endocrine pathways that shape tissue adaptation and disease outcomes. UVB irradiation, in particular, represents a major environmental challenge for epidermal keratinocytes, triggering not only nucleotide excision repair mechanisms but also cytokine release, interferon signaling, and stress-associated pathways that influence cutaneous and systemic homeostasis^1–3.

Prolonged exposure to solar UV radiation primarily leads to skin aging and an increases risk of developing cancer⁴. This is particularly evident in individuals with xeroderma pigmentosum (XP), a genetic disorder characterized by a deficiency in the nucleotide excision repair (NER) pathway, which is a crucial system for correcting UV-induced DNA damage⁵. The process of NER involves a group of more than 40 proteins that work in cascading manner to repair UV-induced DNA lesions, such as 6–4 pyrimidine pyrimidone photoproducts and cyclobutane pyrimidine dimers (CPDs). This process can split into five main enzymatic stages: (i) DNA damage recognition, (ii) Unwinding of the DNA double helix forming a bubble structure, (iii) Verification of the DNA damage followed by the incision of the single stranded DNA harboring the UV lesion at both 5′ and 3′ ends, (iv) Filling the DNA gap via repair replication, utilizing the intact strand as a template, and ultimately, (v) Sealing the newly synthesized DNA patch through ligation⁶. Perturbation in any of the stages of the NER process can lead to the onset of one of the seven Xeroderma Pigmentosum (XP) complementation groups (XP-A to XP-G)⁷. XPC encodes a key protein involved in initiating the GG-NER. Remarkably, mutations in the XPC gene are the most prevalent genetic variation observed in XP. Genodermatosis, also referred to as XP-C disease (OMIM# 278,720)⁸. This protein plays a major role in initiating the GG-NER pathway, by identifying and attaching to DNA helical deformities that occur on the undamaged DNA strand, opposite to the photoproducts produced by UV radiation. XP-C patients show an increased sensitivity to UV light accompanied by an accumulation of DNA damage. This disorder is often termed as skin cancer prone, as patients have a 2,000 to 10,000 folds risk of developing melanomas and non-melanoma skin cancers (NMSCs) at a young age compared to unaffected individuals⁹. It is noteworthy that till date, there is no cure for XP-C syndrome, where the central focus relies on preventative measures using specialized creams against UV exposure¹⁰. Moreover, a detailed understanding of the molecular mechanisms and cellular signaling events underlying this pathological condition is still lacking. To get insight into this, our research group has already established reliable human skin cell models for XP-C disease. In fact, we previously generated, using CRISPR-Cas9 system, XP-C disease models using the three major skin cell types (keratinocytes, fibroblasts, and melanocytes), being immortalized from the human skin, and mimicking XP-C disease in terms of major phenotype characteristics: photosensitivity and absence of photoproducts repair induced by UV exposure¹¹.

Here, and upon using the established XPC KO keratinocytes model, we aimed at deciphering a precise dysregulated molecular signaling pathway linked to XP-C disease. An integrated workflow combining PamGene-based functional phosphoproteomics and MS-based total proteomics was performed., we profiled early signaling events (1-hour post-UVB exposure) and downstream proteomic changes (24 h post-exposure). We identified and upregulated activation of JAK/STAT pathway activation and the downstream type I interferon signaling components in XPC KO keratinocytes under both basal conditions and following UVB exposure. These observations suggest that the constitutive activation of inflammatory signaling pathways may underline the pathophysiology of XP-C, beyond its classical role as a DNA repair disorder.

Results

Selection of a UVB hypersensitive dose based on mortality percentage for upstream and downstream kinase and proteomic analyses

Given that XPC deficiency renders keratinocytes hypersensitive to UVB irradiation^11,12, we designed our experimental conditions to capture early molecular events underlying this phenotype. Accordingly, we selected a UVB dose of 100 J/m² (~ 1 SED) for subsequent analyses. As demonstrated in detail in our previous publication¹¹, at this dose, XPC knockout (KO) N/TERT-2G keratinocytes showed a significantly greater reduction in viability than wild-type (WT) cells, with an approximately 20% difference at 24 h post-irradiation that increased to nearly 50% by 48 h. Analyses were therefore focused at 24 h post-UVB to examine XPC-dependent responses prior to the onset of extensive secondary cytotoxicity.

Importantly, 100 J/m² UVB represents a physiologically relevant, non-lethal exposure for human epidermal keratinocytes, corresponding to approximately one standard erythema dose (SED)¹³. For molecular profiling, we analyzed early signaling responses at 1 h post-UVB and downstream molecular changes at 24 h, enabling investigation of both immediate and delayed mechanisms contributing to UVB hypersensitivity in XPC-deficient keratinocytes.

This schematic illustrates the two-stage, time-resolved experimental design of the study. Upper panel: Early signaling events were analyzed 1 h post-UVB irradiation using PamGene-based phosphotyrosine kinase profiling in wild-type and XPC knockout keratinocytes. Lower panel: Downstream cellular responses were analyzed 24 h post-UVB irradiation using TMT-based quantitative LC–MS/MS proteomics. Results from both workflows were independently analyzed and subsequently integrated using protein–protein interaction and pathway enrichment analyses, leading to the identification of JAK/STAT-mediated type I interferon signaling as a dysregulated pathway in XPC-deficient keratinocytes. Right panel: Key pathway components were validated by immunoblotting.

Aberrant phosphorylation of the JAK/STAT signaling components in XPC knockout keratinocytes following 1 h of UVB irradiation

Observing that XPC KO cells exhibit increasing hypersensitivity to UVB irradiation over time prompted us to elucidate the molecular and signaling events underlying this phenomenon. Our workflow strategy integrated two approaches (Fig. 1). In the first we used PamGene-based functional phosphoproteomics analysis to identify activated phosphorylation sites and predict potential kinases while the second we employed total proteomics analysis to unravel deregulated signaling cascades.

Fig. 1.

A novel strategy for identifying dysregulated signaling pathways in XPC knockout cells upon UVB irradiation.

Pamgene kinase activity profiling was applied to evaluate the overall activity of protein tyrosine kinases (PTKs). Kinomic profiles were assessed in XPC KO keratinocytes versus wild-type control cells at basal state as well as 60 min following UVB irradiation respecting the same experimental conditions and relying on our keratinocytes dose response curve from our previous work¹¹. Following the quantification of the signals for the four conditions, it was crucial to decipher and quantify their dysregulated phosphosites. Following identification, quantification, and quality control filtering, this procedure unraveled 195 identified PTK phosphosites which passed QC test (excluding background system effects) we then generated (Fig. 2).

Fig. 2.

Differential phosphorylation signatures in wild-type (control) and XPC KO N/TERT-2G keratinocyte cells following 1 h after UVB Irradiation.

The phosphorylation profiles varied significantly between the XPC KO cells and wild type (control) cells under the experimental conditions examined (Fig. 3A and B). For instance, 64 PTK phosphosites were downregulated in UVB irradiated versus non irradiated wild type control cells (WT IR vs. WT NIR). On the other hand, 94 PTK phosphosites were upregulated in UVB irradiated versus non irradiated XPC KO cells (XPC KO IR vs. XPC KO NIR). Interestingly, 104 PTK phosphosites were upregulated in UVB-irradiated XPC KO cells versus UVB-irradiated wild-type cells (XPC KO IR vs. WT IR). These observations suggested a potential role for XPC in regulating the phosphorylation status of PTKs.

Fig. 3.

Alterations in PTK phosphorylation dynamics following 1 h of UVB irradiation in wild-type (control) and XPC KO N/TERT-2G Cells. (A) Table showing the number of significantly altered PTK phosphosites across various conditions (WT IR vs. WT NIR, XPC KO IR vs. XPC KO NIR, XPC KO IR vs. WT IR). (B) Volcano plots illustrating the distribution of differentially phosphorylated PTK phosphosites in response to UVB irradiation in various conditions. Comparisons shown: WT IR vs. WT NIR, XPC KO IR vs. XPC KO NIR, and XPC KO IR vs. WT IR. Red points indicate significantly altered phosphorylation (p < 0.05). (C1). Heatmap presentation of differentially altered PTK peptide phosphosites in XPC KO keratinocytes at basal state versus UVB irradiation. This heatmap compares XPC KO non-UVB irradiated versus wild type non-UVB irradiated keratinocytes (noted as XPC KO NIR) along with XPC KO UVB irradiated normalized to wild type UVB irradiated keratinocytes (noted as XPC KO IR). The results presented are the mean of three biological replicates (N = 3). (C2). Heatmap presentation of differentially altered PTK peptide phosphosites in wild-type versus XPC KO keratinocytes following UVB irradiation. This heatmap compares both wild-type UVB irradiated versus non-UVB irradiated keratinocytes (noted as WT) along with XPC KO UVB irradiated versus non-UVB irradiated keratinocytes (noted as XPC KO) for PTK peptide phosphosites. JAK/STAT family: JAK1, JAK2, TYK2, JAK3, and STAT4 (in red) were further selected for future validation experiments. The results presented are the mean of three biological replicates (N = 3). JAK/STAT family: JAK1, JAK2, JAK3, and STAT4 (in red) were further selected for future validation experiments. The logFC ranged from + 3 to − 3 for significant peptide phosphorylation sites for PTK. The results presented are the mean of three biological replicates (N = 3).

Heatmap showing differential tyrosine phosphorylation profiles across wild-type (control) and XPC knockout (KO) cell lines under non-irradiated and irradiated conditions. Phosphopeptide intensities are represented as a gradient from low (blue) to high (red), highlighting condition-specific and genotype-specific signaling responses. Biological triplicates (N = 3) are shown for each group. Key phosphosites are annotated on the y-axis, and samples are grouped along the x-axis according to treatment and genotype.

In a further step, we designed a heatmap comparing phosphosites in IR/NIR wild-type control cells versus IR/NIR XPC KO cells (Fig. 3C1) as well as another heatmap comparing phosphosites in NIR XPC KO /NIR wild-type cells versus IR XPC KO /IR wild-type cells (Fig. 3C2). Figure 3C1 revealed a significant upregulation in the phosphorylation status across a broad range of PTK peptide phosphosites following XPC knockout. Among the identified phosphosites, were those of the JAK/STAT family: JAK1, JAK2, TYK2, JAK3, as well as STAT4. Interestingly, the phosphorylation status of those members of the JAK/STAT signaling pathway were also upregulated in response to UVB irradiation as monitored in IR- versus NIR-XPC KO cells (3C2). These observations highlight a potential role for the deregulated JAK/STAT signaling pathway in mediating the XPC KO phenotype in response to UVB irradiation.

The upregulation of these players reflects the robustness of the dysregulation of this pathway and can exclude the false positive effect. Thus, JAK/STAT signaling pathway seemed to be a target of dysregulation for XPC KO keratinocytes following UVB irradiation. The phosphosite analysis showed the dysregulation of JAK/STAT family members in XPC KO keratinocytes post-UVB irradiation.

Global proteomic profiling reveals activation of type I interferon downstream effectors 24 h After UVB irradiation in XPC-deficient keratinocytes

Protein samples derived from non-irradiated wild-type cells, UVB-irradiated wild-type cells, non-irradiated XPC KO cells and UVB-irradiated XPC KO cells were prepared then analyzed by mass spectrometry as mentioned in the materials and methods section.

A total of 6699 proteins were reliably identified. Statistical analyses were then carried out to identify the proteins that were differentially abundant between the conditions compared in pairs. 390 proteins were identified to be differentially abundant in wild-type keratinocytes following UVB irradiation compared to non-irradiated ones (Fig. 4A). When comparing XPC KO UVB irradiated and non-UVB irradiated keratinocytes.

Fig. 4.

Volcano plot representation of differentially expressed proteins across keratinocyte genotypes and UVB exposure conditions. Volcano plots illustrate the significantly upregulated and downregulated proteins identified in three pairwise comparisons involving N/TERT-2G wild-type and XPC knockout (KO) keratinocytes, under basal conditions and following UVB irradiation. (A) Wild-type keratinocytes: UVB-irradiated versus non-irradiated (WT IR vs. WT NIR). (B) XPC KO keratinocytes: UVB-irradiated versus non-irradiated (XPC IR vs. XPC NIR). (C) UVB-irradiated XPC KO versus UVB-irradiated wild-type keratinocytes (XPC IR vs. WT IR). Thresholds applied across all comparisons were log₂(fold change) ≥ ± 0.3 (corresponding to ≥ 1.23-fold change) and LIMMA p-value ≤ 0.01. False discovery rates (FDRs) were controlled for each comparison: WT UVB vs. non-UVB (1.14%), XPC KO UVB vs. non-UVB (1.08%), XPC KO vs. WT at baseline (1.34%), and XPC KO vs. WT post-UVB (1.18%). Volcano plots were generated using GraphPad Prism v8, with log₂ (fold change) on the x-axis and –log₁₀(p-value) on the y-axis to visualize the distribution and significance of differentially expressed proteins.

(XPC KO IR vs. XPC KO NIR), 791 proteins were identified to be significantly dysregulated (upregulated or downregulated), representing more than 10% of the quantified proteins (Fig. 4B). When comparing XPC KO UVB irradiated vs. wild type UVB irradiated keratinocytes (XPC KO IR vs. WT IR), 413 proteins were found to be differentially abundant (upregulated or downregulated) (Fig. 4C). To more effectively illustrate the dysregulated proteins, a heatmap was generated, highlighting the highly upregulated proteins (log2foldchange) which were set on the top list whereas the most downregulated proteins were set on the bottom list. It is noteworthy that IR XPC KO normalized to IR-wild-type was used as a reference. The top list proteins included Interferon-induced GTP-binding protein (MX1), Serpin Family B Member 3 (SERPINB3), S100 calcium-binding protein A12 (S100A12), Interferon Induced Protein With Tetratricopeptide Repeats 2 (IFIT-2), Interferon Induced Protein With Tetratricopeptide Repeats 3 (IFIT-3), Interferon Induced Protein With Tetratricopeptide Repeats 1 (IFIT-1), Interferon-induced GTP-binding protein (MX2), Suprabasin (SBSN), Interleukin 1 beta (IL-1β), S100 calcium-binding protein A7 (S100A7), Calmodulin Like 5 (CALML5), 2′-5′-Oligoadenylate Synthetase 1 (OAS1), Interferon Regulatory Factor 9 (IRF9), and Interferon-stimulated gene 15 (ISG15) (Fig. 5). The STRING protein–protein interaction network software (Fig. 6A) was utilized to analyze the top cluster of dysregulated proteins together with members of the JAK/STAT family (JAK1, JAK2, JAK3, STAT1, STAT2, and STAT3). Although these JAK/STAT proteins were identified as deregulated in the proteomics analysis, they were not among the top-ranked proteins; thus, they were included to assess whether they form an interaction network. Interestingly, the results showed a robust protein-protein interaction network between this cluster of dysregulated proteins and the JAK/STAT family members, except for suprabasin, which was not linked to the network.

Fig. 5.

Heat map presentation of the significantly and differentially upregulated and downregulated proteins in four various conditions encompassing N/TERT-2G wild-type and XPC KO keratinocytes at basal state and following UVB irradiation. This heatmap presentation shows the differential expression profile of 413 proteins (arranged from the highly upregulated on the toplist to the downregulated on the bottom list according to the condition XPC KO IR vs. WT IR used as a reference) and top 14 upregulated list in XPC KO IR conditions were mentioned here. Each row is a protein and each column corresponds to a specific condition starting from XPC KO UVB irradiated vs. wild-type UVB irradiated (XPC KO IR vs. WT IR), XPC KO UVB irradiated vs. non-UVB irradiated (XPC KO IR vs. XPC KO NIR), XPC KO non-UVB irradiated vs. wild-type non-UVB irradiated (XPC KO NIR vs. WT NIR), and wild-type UVB irradiated vs. non-UVB irradiated (WT IR vs. WT NIR). The 14 top list proteins selected according to the condition XPC KO IR vs. WT IR compared to the other three conditions were Interferon-induced GTP-binding protein (MX1), Serpin Family B Member 3 (SERPINB3), S100 calcium-binding protein A12 (S100A12), Interferon Induced Protein With Tetratricopeptide Repeats 2 (IFIT-2), Interferon Induced Protein With Tetratricopeptide Repeats 3 (IFIT-3), Interferon Induced Protein With Tetratricopeptide Repeats 1 (IFIT-1), Interferon-induced GTP-binding protein (MX2), Suprabasin (SBSN), Interleukin 1 beta (IL-1β), S100 calcium-binding protein A7 (S100A7), Calmodulin Like 5 (CALML5), 2′-5′-Oligoadenylate Synthetase 1 (OAS1), Interferon Regulatory Factor 9 (IRF9), and Interferon-stimulated gene 15 (ISG15). The FunRich Version 3.1.4 was used to construct this heatmap. The Row Z score fluctuated from − 1 to 1.

Fig. 6.

Integrated protein-protein interaction network and enriched biological pathways of dysregulated proteins in XPC KO UVB-irradiated cells compared to wild-type UVB-irradiated cells, highlighting JAK/STAT signaling. (A) This figure presents an integrated view of the top dysregulated proteins (upregulated) in XPC knockout (KO) cells following UVB irradiation compared to wild-type (WT) UVB-irradiated cells. The protein-protein interaction (PPI) network, constructed using STRING database v11.5, reveals extensive associations among 13 key upregulated proteins—MX1, MX2, SERPINB3, S100A12, IFIT1, IFIT2, IFIT3, IL-1β, S100A7, CALML5, OAS1, IRF9, and ISG15—along with JAK/STAT family members (JAK1, JAK2, JAK3, STAT1, STAT2, STAT3), previously identified as a dysregulated upstream kinase cluster. The network underscores the central role of immune and interferon-related signaling. (B) In parallel, enrichment analysis using g: Profiler v2.1 revealed significant clustering of these proteins into common biological pathways based on Reactome database annotations. The most prominent pathways include Type I interferon-alpha/beta signaling, general interferon signaling, and cytokine signaling in the immune system. Together, these findings suggest a strong immune and interferon-driven response in XPC KO cells upon UVB exposure, mediated in part by JAK/STAT signaling. STRING analysis was used exclusively as a hypothesis-generating and visualization tool to contextualize dysregulated proteins within known signaling networks and was not intended to infer novel physical protein–protein interactions.

To better predict the signaling pathway in which this protein network is involved, g:Profiler software was applied (Fig. 6B).

Our results unraveled Type I Interferon alpha/beta signaling pathway to be on the top list including MX1, MX2, IFIT1, IFIT2, IFIT3, JAK1, STAT1, STAT2, OAS1, IRF9, and ISG15. The second top hit was Interferon signaling including MX1, MX2, IFIT1, IFIT2, IFIT3, JAK1, JAK2, STAT1, STAT2, OAS1, IRF9, and ISG15. The third top hit was Cytokine Signaling in Immune system including MX1, MX2, IFIT1, IFIT2, IFIT3, JAK1, JAK2, JAK3, STAT1, STAT2, STAT3, S100A12, IL1β, OAS1, IRF9, and ISG15.

Validation of upstream and downstream components of type I interferon signaling through the JAK/STAT pathway in XPC knockout keratinocytes

The above observations indicated an upregulated role for JAK/STAT signaling pathway in UVB-irradiated XPC KO keratinocytes. Accordingly, a set of experiments were conducted to validate this observation.

In a first step, we assessed the phosphorylation/expression status of several STAT proteins including STAT1, STAT2 and STAT3. Western blot analysis was therefore performed using antibodies against alpha phosphoTyr-701-STAT-1 (91 kDa), phosphoTyr-690-STAT-2 (97 kDa), and phosphoTyr-705-STAT-3 (86 kDa) (Fig. 7A). In the case of alpha phosphoTyr-701-STAT-1 (91 kDa), the expression/phosphorylation level was strikingly higher in UVB-irradiated XPC KO cells versus UVB-irradiated wild-type control cells (Fig. 7B). Moreover, the expression status was significantly higher in UVB-irradiated XPC KO cells versus non-irradiated XPC KO cells (Fig. 7B). Similarly in the case of phosphoTyr-690-STAT-2 (97 kDa), UVB-irradiated XPC KO cells showed the highest expression/phosphorylation level. This was 5x higher than that of UVB-irradiated wild-type cells, 2.5x more than that exhibited by non-irradiated wild-type cells, and 1.11x greater than that of non-irradiated XPC KO cells (Fig. 7C). Further, in the case of phosphoTyr-705-STAT-3 (86 kDa), the expression/phosphorylation levels were substantially higher in XPC KO cells versus wild-type control cells in absence or presence of UVB irradiation (Fig. 7D).

Fig. 7.

Increased phosphorylation of STAT proteins in XPC knockout keratinocytes under basal conditions and after UVB exposure. (A) Western blot analysis showing increased phosphorylation of STAT family members specifically phospho-Tyr701-STAT1 (91 kDa), phospho-Tyr690-STAT2 (97 kDa), and phospho-Tyr705-STAT3 (86 kDa) in XPC-deficient keratinocytes compared to wild-type cells, both at baseline and post-UVB irradiation. (B–D) Quantification of fold changes in expression levels for (B) phospho-Tyr701-STAT1, (C) phospho-Tyr690-STAT2, and (D) phospho-Tyr705-STAT3. Protein lysates were resolved via SDS-PAGE and transferred to nitrocellulose membranes. Detection was performed using phospho-specific antibodies, with GAPDH (37 kDa) serving as a loading control. Western blot quantification was normalized to GAPDH, and fold change was calculated relative to the XPC KO IR condition (set to 1). Statistical significance was assessed using unpaired t-test *p-value < 0.05, **p-value < 0.01, ***p-value < 0.001, ****p-value < 0.0001. Data represent mean values from three independent replicates (N = 3). Uncropped blots are shown in Figure S1.

In a second step, western blot analysis was used to assess the expression levels of several downstream effectors (MX1, MX2, IFIT1, IFIT3, IRF9, and ISG15) of the JAK/STAT signaling pathway (Fig. 8). Interestingly, the expression levels of IRF-9 (Fig. 8B), IFIT1 (Fig. 8C), IFIT3 (Fig. 8D), MX1 (Fig. 8F), ISG15 (Fig. 8H) and MX2 (Fig. 8I) were strikingly higher in XPC KO cells versus wild-type cells either in absence or presence of UVB irradiation. Intriguingly, the expression levels of those genes were significantly upregulated following exposure of XPC KO cells to UVB-irradiation (Fig. 8). For IRF-9 (Fig. 8A and B), the results showed a significant upregulation of IRF-9 in XPC KO UVB irradiated keratinocytes (5.5 folds), in XPC KO non-UVB irradiated keratinocytes (3.3 folds), and a slight increase for wild-type UVB irradiated keratinocytes (1.5 folds) compared to the reference utilized which was wild-type non-UVB irradiated keratinocytes set at a constant 1-fold to compare. For IFIT1 (Figs. 8C and 9E), the results showed a significant upregulation of IFIT1 in XPC KO UVB irradiated keratinocytes (15.5 folds), in XPC KO non-UVB irradiated keratinocytes (7.5 folds), and a constant expression level for wild-type UVB irradiated keratinocytes (1-fold) compared to the reference utilized which was wild-type non-UVB irradiated keratinocytes set at a constant 1 fold to compare. For IFIT3 (Fig. 8D and E), the results showed a significant upregulation of IFIT3 in XPC KO UVB irradiated keratinocytes (17.5 folds), in XPC KO non-UVB irradiated keratinocytes (5 folds), and a slight increase for wild-type UVB irradiated keratinocytes (1.75 folds) compared to the reference utilized which was wild-type non-UVB irradiated keratinocytes set at a constant 1-fold to compare. For MX1 (Fig. 8F and E), the results showed a significant upregulation of MX1 in XPC KO UVB irradiated keratinocytes (15.7 folds), in XPC KO non-UVB irradiated keratinocytes (5.5 folds), and a constant expression level for wild-type UVB irradiated keratinocytes (1-fold) compared to the reference utilized which was wild-type non-UVB irradiated keratinocytes set at a constant 1 fold to compare. For MX2 (Fig. 8G and I), the results showed a significant upregulation of MX2 in XPC KO UVB irradiated keratinocytes (20.5 folds), in XPC KO non-UVB irradiated keratinocytes (9 folds), and a slight downregulation for wild type UVB irradiated keratinocytes (0.75 folds) compared to the reference utilized which was wild type non-UVB irradiated keratinocytes set at a constant 1-fold to compare. For ISG15 (15 kDa) (Fig. 8H and G), the results showed a significant upregulation of ISG15 in XPC KO UVB irradiated keratinocytes (19 folds), in XPC KO non-UVB irradiated keratinocytes (2 folds), and a constant expression level for wild-type UVB irradiated keratinocytes (1-fold) compared to the reference utilized which was wild-type non-UVB irradiated keratinocytes set at a constant 1-fold to compare.

Fig. 8.

UVB irradiation induces robust activation of Type I interferon downstream effectors in XPC-deficient keratinocytes. Immunoblotting and quantitative analysis reveal elevated expression of Type I interferon downstream effectors in XPC knockout (KO) keratinocytes compared to wild-type (WT) controls following UVB irradiation. (A, E, G) Immunoblots showing expression of IRF9 (48 kDa), IFIT1 (56 kDa), IFIT3 (55 kDa), MX1 (76 kDa), MX2 (80 kDa), and ISG15 (15 kDa) across four experimental groups: XPC KO irradiated (IR), XPC KO non-irradiated (NIR), WT IR, and WT NIR. GAPDH (37 kDa) served as loading control. (B–D, F, H–I) Quantification of fold-change in protein expression for IRF9 (B), IFIT1 (C), IFIT3 (D), MX1 (F), ISG15 (H), and MX2 (I) reveals significantly higher induction in XPC KO cells 24 h post-UVB exposure compared to other groups. Western blot quantification was normalized to GAPDH, and fold change was calculated relative to the WT non-NIR condition (set to 1). Data represent mean values from three independent replicates (N = 3). Statistical analysis performed using using unpaired t-test *p-value < 0.05, **p-value < 0.01, ***p-value < 0.001, ****p-value < 0.0001. Uncropped blots are shown in Figure S1 Altogether. these observations are in support of a JAK/STAT mediated-upregulation of Type I interferon alpha/beta signaling in XPC KO keratinocytes at basal state as well as following UVB irradiation.

Fig. 9.

Hypothetical model of Type I interferon pathway activation via JAK/STAT signaling in XPC-deficient keratinocytes. Schematic representation of a proposed mechanism by which Type I IFN signaling is selectively activated in XPC knockout (KO) keratinocytes. Type I interferons signal through the heterodimeric IFNAR1/IFNAR2 receptor complex, which undergoes conformational changes upon ligand binding, initiating phosphorylation of Tyk2 and JAK1. This leads to the activation of STAT1 and STAT2, and occasionally STAT3 via JAK kinases. Activated STATs dimerize to form transcriptionally active complexes. The STAT1–STAT2 heterodimer associates with IRF9 to form ISGF3, which translocates to the nucleus and binds to interferon-stimulated response elements (ISREs), inducing expression of pro-inflammatory genes such as MX1, MX2, IFIT1, IFIT3, and ISG15. STAT1 and STAT3 homodimers can also independently bind to gamma-activated sequence (GAS) elements to modulate similar inflammatory responses.

Discussion

This study was designed as a discovery-driven analysis aimed at identifying signaling pathways that are dysregulated as a consequence of XPC loss, rather than as a functional interrogation of individual pathways. By integrating early phosphotyrosine kinase activity profiling, global quantitative proteomics, and targeted biochemical validation, we define a robust and reproducible molecular signaling signature associated with XPC deficiency in human keratinocytes. Using this multi-layered approach, we identify a consistent upregulation of the JAK/STAT pathway and its downstream type I interferon–related effectors under both basal conditions and following UVB exposure. These findings delineate a disease-associated inflammatory signaling landscape without implying direct cause-and-effect relationships, and provide a foundation for future mechanistic studies aimed at dissecting the functional contribution of interferon signaling to XP-C pathology.

XP-C disease is one of the most prevalent subtypes of xeroderma pigmentosum¹⁴ and is typically caused by nonsense mutations, frameshifts, or deletions in the XPC gene¹⁵. These mutations often lead to complete loss of the XPC protein and negligible mRNA expression¹⁵. XP-C cells are profoundly hypersensitive to UVB and exhibit impaired repair of UV-induced DNA lesions such as 6−4 photoproducts (6-4PPs) and cyclobutane pyrimidine dimers (CPDs), due to a defective global genome nucleotide excision repair (GG-NER) mechanism¹². Ultraviolet B (UVB) radiation exerts its biological effects through multiple, interconnected mechanisms that extend beyond the direct induction of DNA photolesions. In addition to generating cyclobutane pyrimidine dimers and 6–4 photoproducts, UVB exposure promotes oxidative stress through the production of reactive oxygen species, leading to indirect DNA damage, lipid peroxidation, and protein oxidation¹⁶. UVB can also activate a broad range of cellular signaling pathways, including MAPK, NF-κB, and JAK–STAT signaling, thereby modulating inflammatory responses, apoptosis, cell cycle regulation, and immune surveillance¹⁷. Furthermore, UVB has been shown to alter chromatin structure, transcriptional programs, and cytokine secretion, contributing to long-term changes in cellular homeostasis¹⁸. The convergence of direct genotoxic stress and signaling-mediated responses underscores the complexity of UVB-induced cellular damage, particularly in contexts where DNA repair pathways are compromised². Given that the epidermis is the primary site of UV exposure, our experimental model was based on keratinocytes: the most abundant epidermal cells and a key target in non-melanoma skin cancers (NMSCs), which occur at rates up to 10,000 times higher in XP-C patients than in the general population.

Despite the clinical importance of keratinocytes in XP-C pathology, the field has lacked reliable experimental models due to the rarity of patient-derived keratinocyte cells and challenges associated with their limited lifespan and absence of genetically isogenic controls¹⁹. This does not exclude the importance of validating our results on XP-C keratinocytes from patients in the future, once such samples become accessible. Nevertheless, our current strategy provides a valuable platform to facilitate the elucidation of molecular signaling pathways, potentially enabling direct validation and therapeutic development in more advanced models. Additionally, mouse models can offer partial data but are limited by significant interspecies differences in skin biology and difficulties in maintaining primary mouse keratinocytes in vitro as there is no reliable media and protocols to culture them as well as maintain them in culture²⁰. To overcome these limitations, we sought to choose the immortalized human N/TERT-2G keratinocyte cell line. This model has a normal karyotype²¹, supports long-term culture²², and faithfully recapitulates XP-C phenotypes, including the hypersenisitve phenotype towards UVB irradiation and delayed repair of UVB-induced DNA damage¹¹. Its utility for therapeutic research has also been demonstrated. Kobaisi et al. showed that targeting PIK3C3 and LATS1 led to over 20% of repair restoration of 6-4PP DNA damage using this XPC KO model²³.

To dissect the molecular alterations in XPC-deficient keratinocytes, we implemented a dual-stage proteomics workflow. Phosphoproteomics was used to identify early signaling changes one-hour post-UVB irradiation, while total proteome analysis at 24 h revealed broader downstream effects. This integrated approach enabled the construction of a high-resolution signaling map, revealing that type I IFN signaling was aberrantly activated in XPC KO cells both at baseline and after UVB exposure through the JAK/STAT axis. It’s important to mention that we observed variability in PTK phosphosite intensities across biological replicates, particularly in wild-type and XPC KO keratinocytes as shown in Fig. 2. This likely reflects both biological and technical factors: early kinase signaling is highly dynamic, with rapid phosphorylation–dephosphorylation cycles, and UVB-induced stress responses can vary even among clonal cells. Technically, PamGene assays measure functional phosphorylation capacity rather than static phosphosite abundance, so modest differences in extraction efficiency or kinase accessibility can increase signal dispersion. Importantly, stringent quality control, batch correction, and kinetic filtering ensured that only high-confidence phosphosites were analyzed. Moreover, pathway-level conclusions such as JAK/STAT activation are robust, supported by coherent kinase-family enrichment, total proteomics, and independent western blot validation. Thus, variability at individual phosphosites does not undermine our main conclusions regarding XPC deficiency and UVB-amplified type I interferon signaling. This strategy not only illuminates the cellular pathways affected by UVB exposure but also represents a powerful approach for uncovering disease-related mechanisms on a broader scale. By mapping critical signaling disruptions, it has the potential to inform the development of targeted therapies and advance our understanding of complex disease biology.

A major strength of this study was that we combined phosphoproteomics and total proteomics analysis. Phosphoproteomics was used to identify early signaling changes one-hour post-UVB irradiation, while total proteome analysis revealed broader downstream effectors. Although our phosphoproteomic analyses revealed strong dysregulation of several signaling nodes, including EGFR, CRK, and CTNNB1, which are classically associated with receptor tyrosine kinase and Wnt/adhesion pathways, pathway prioritization in this study was not based on individual fold-change magnitude. Instead, we focused on the JAK/STAT pathway because it uniquely exhibited coordinated activation across multiple experimental layers. Specifically, JAK/STAT signaling showed early and concurrent phosphorylation of several pathway components at the kinase level, followed by a sustained downstream response characterized by robust induction of canonical type I interferon-stimulated genes. Importantly, this pathway-level activation was independently validated by biochemical analyses at both phosphorylation and protein expression levels. While RTK- and adhesion-related pathways are clearly perturbed in XPC-deficient keratinocytes and may contribute to UVB-induced stress responses, the JAK/STAT axis emerged as a convergent and functionally amplified signaling cascade linking XPC loss to a persistent type I interferon-driven inflammatory state.

Type I interferon signaling plays a central role in the cellular response to DNA damage, viral infection, and immune regulation. This pathway is primarily mediated by interferon-α and interferon-β, which signal through the interferon-α/β receptor (IFNAR) to activate the JAK–STAT cascade, leading to the transcription of interferon-stimulated genes (ISGs). Beyond its antiviral functions, type I interferon signaling has been implicated in the regulation of cell cycle progression, apoptosis, and DNA damage responses, including nucleotide excision repair, and its dysregulation has been associated with chronic inflammation, autoimmunity, and cancer²⁴. In line with these observations, the JAK/STAT pathway (Fig. 9) is activated upon cytokine binding to IFNAR1/IFNAR2 receptor complexes, which triggers phosphorylation of TYK2 and JAK1, followed by downstream activation of STAT1, STAT2, and STAT3²⁵. Upon UVB exposure, we found a strong induction of pSTAT1 and a modest upregulation of pSTAT2, indicating activation of the type I interferon signaling pathway. Furthermore, pSTAT3 was already elevated under basal conditions in XPC KO keratinocytes, suggesting a constitutive dysregulation. We further confirmed the phosphorylation of specific tyrosine residues: Tyr701 on STAT1α and Tyr690 on STAT2, which are known to be phosphorylated by JAK1 and TYK2, triggering downstream interferon responses^26,27. Additionally, STAT3 activation involves phosphorylation at Tyr705, and is known to crosstalk with JAK pathway²⁸. By quantifying the phosphorylation status of these residues, we confirmed the validity of our hypothesis. These phosphorylated STATs form homodimers or heterodimers and translocate to the nucleus, driving transcription of interferon-stimulated genes (ISGs) via binding to GAS or ISRE elements. In our model, activation of this cascade was accompanied by upregulation of hallmark ISGs such as ISG15, MX1, MX2, IFIT1, and IFIT3, suggesting a persistent inflammatory state in XPC-deficient keratinocytes²⁵.

While the type I interferon pathway emerged as a prominent feature of the dysregulated signaling landscape in XPC knockout keratinocytes, direct pharmacological suppression of this pathway is unlikely to be straightforward in this setting. XPC loss combined with UVB exposure induces broad perturbations across interconnected inflammatory signaling networks, including TLR, NF-κB, SRC, MAPK, and c-MET pathways, as revealed by our phosphoproteomic and global proteomic analyses. This extensive signaling crosstalk complicates attempts to selectively inhibit the interferon response without concomitantly affecting parallel pathways.

Within this complex signaling context, our data nevertheless highlight the JAK/STAT axis as a prioritized and experimentally tractable node within the broader interferon-associated network. This prioritization is supported by (i) the magnitude and consistency of JAK/STAT activation across independent phosphoproteomic, proteomic, and biochemical validation layers, (ii) the central positioning of JAK/STAT components within the inferred protein–protein interaction network, and (iii) parallels with inflammatory skin disorders in which dysregulated type I interferon signaling and clinical responses to JAK inhibition have been described. Importantly, these observations do not imply therapeutic efficacy or establish causality; rather, they provide a rationale for focused functional interrogation. Unbiased genetic approaches, such as high-throughput loss-of-function screens targeting JAK/STAT and inflammation-related genes, will be required to determine whether modulation of this axis can influence UV hypersensitivity or other disease-relevant phenotypes associated with XPC deficiency.

Interestingly, this IFN-driven transcriptional signature mirrors that seen in other UV-related inflammatory skin disorders, including cutaneous lupus erythematosus²⁹, psoriasis³⁰, and vitiligo^31,32, and morphea^33,34, where type I IFN activation contributes to tissue damage and immune dysregulation. These similarities raise the possibility that XP-C could share molecular features with inflammatory skin diseases, extending beyond its classification as a DNA repair disorder. Thus, our findings not only illuminate the contribution of the type I IFN pathway to XP-C pathogenesis but also position this disease within a broader inflammatory framework. Whether the chronic activation of type I interferon signaling directly impacts nucleotide excision repair efficiency represents an important but distinct question that warrants future mechanistic investigation.

Recent studies indicate that XPC can impact tumor-suppressor pathways beyond its classic role in nucleotide excision repair. For instance, XPC loss has been shown to promote melanomagenesis through downregulation of CDKN2A in melanocytes. In our keratinocyte-based proteomic analyses, however, CDKN2A was either undetected or not significantly regulated under basal or UVB-irradiated conditions. This likely reflects cell-type–specific effects of XPC deficiency, as melanocytes and keratinocytes differ in cell-cycle control, UV response, and oncogenic susceptibility. While CDKN2A deregulation may drive XPC-mediated melanomagenesis, our data suggest that in keratinocytes, XPC loss primarily activates inflammatory and type I interferon pathways. These findings underscore the pleiotropic and context-dependent roles of XPC in skin homeostasis and cancer. Further research is required to fully comprehend the precise mechanisms and implications of this dysregulation in XPC KO keratinocytes and its association with skin inflammatory disorders.

Conclusion

Xeroderma pigmentosum, complementation group C (XP-C), is a rare autosomal recessive disorder resulting from mutations in the XPC gene, leading to impaired nucleotide excision repair and marked sensitivity to ultraviolet radiation. Affected individuals face a significantly elevated risk of early-onset skin cancers and require lifelong photoprotection and vigilant dermatologic monitoring. Despite the absence of curative therapies, understanding the molecular consequences of XPC loss remains crucial. This study provides the first evidence of JAK/STAT and type I IFN signaling hyperactivation in XPC KO keratinocytes, offering a remarkable breakthrough into better understanding the molecular mechanisms underlying the disease’s clinical manifestations. Further investigations into the interplay between defects in DNA repair machinery and deregulated inflammatory responses might open new avenues for targeted therapeutic procedures in XPC and other related genetic disorders.

Materials and methods

Cell lines

The human immortalized male epidermal keratinocyte cell line (N/TERT-2G) was generously provided by Dr. James Rheinwald’s laboratory at Harvard Medical School, Boston, USA²². The N/TERT-2G cells electroporated with a scrambled sgRNA (negative control) marked as wild type (WT) and the XPC KO cells were cultured in EpiLife medium (60 µM calcium; Gibco™, cat. #MEPI500CA) supplemented with human keratinocyte growth supplement (HKGS, Gibco™, cat. #S0015), containing 0.01 µg/ml recombinant human insulin-like growth factor I, 0.2% (v/v) bovine pituitary extract (BPE), 5 µg/ml bovine transferrin, 0.18 µg/ml hydrocortisone, and 0.2 ng/ml human epidermal growth factor. Additionally, CaCl₂ (340 µM; Sigma-Aldrich, Saint Louis, USA) and 1% penicillin/streptomycin were added. Cells were washed with 8 mL phosphate-buffered saline (PBS, pH 7.4; Gibco™) before dissociation from the T75 cm² culture flask using 3 mL of 0.05% trypsin/EDTA (Gibco™) for 5–10 min at 37 °C, depending on the cell line. Trypsinization was halted by adding 8 mL of complete culture medium. The cells were then centrifuged at 100× g for 5 min, and the supernatant was discarded.

UVB dose choice

The photosensitivity of keratinocytes (XPC KO vs. WT) was assessed following escalating UVB doses, as previously published in our earlier work¹¹. For the current study, we selected specific UVB dose based on reanalysis of the previously collected data points to focus on the range that best distinguishes the responses of XPC KO and WT cells. Cell viability at 24- and 48-hours post-irradiation was measured using a trypan blue exclusion assay (ThermoFisher Scientific, MA, USA) and normalized to non-irradiated controls (0 J/cm², set at 100%). Experiments were performed in triplicate (N = 3).

Kinase phosphosites profiling using pamgene technology

Four samples were prepared in triplicates to decipher the kinase activity signature of wild-type and XPC KO N/TERT-2G keratinocytes at basal state and after UVB irradiation (N = 3). The four samples were wild-type non-irradiated, wild type irradiated, XPC KO non-irradiated, and XPC KO irradiated. Cells were seeded in 6 well plates to reach 80% of confluence. Afterwards, they were either irradiated or not and maintained for 1 h at 37 °C. Cells from all experimental conditions were trypsinized and lysed using M-PER™ Mammalian Extraction Buffer (Cat. #78503, Thermo Fisher Scientific, USA) supplemented with Halt™ Protease Inhibitor Cocktail, EDTA-free (100x) (Cat. #78437, Thermo Fisher Scientific, USA) and Halt™ Phosphatase Inhibitor Cocktail (100x) (Cat. #78428, Thermo Fisher Scientific, USA). Protein concentration was determined using the bicinchoninic acid (BCA) or Smith assay, both of which exhibit high sensitivity toward proteins with minimal interference from other substances. Phosphorylation activity was assessed using PamChip^® microarrays (PamGene, Cat. #32501), which contain immobilized peptides. Protein samples were prepared by diluting extracted proteins with M-PER buffer supplemented with protease and phosphatase inhibitor cocktails to final concentrations of 5 µg for protein tyrosine kinase assays per condition. The assay protocol followed manufacturer instructions (PamGene International B.V.) as described in the provided protocol documents (https://pamgene.com/wp-content/uploads/2020/09/PTK-assay-protocol-PS12-vs-3-2019-01.pdf). PamChip^® microarrays were loaded into the PamStation^®12 (PamGene International B.V.), and chips were blocked with 1% bovine serum albumin (BSA) before sample application. A reaction mix containing diluted primary antibodies was added immediately before sample loading. Following a one-and-a-half-hour incubation period, a detection mix was applied to the chips. Fluorescently labeled anti-phospho-antibodies were used to detect tyrosine phosphorylation activity. The PamStation^®12 recorded images of the arrays using a CCD camera. Experiments were performed in triplicate (N = 3).

Kinase phosphosites profiling: quality control and normalization

A flag system was used to indicate quality control (QC) levels based on three criteria: Phosphosite signal strength (99th percentile of S100 arbitrary units, AU), Phosphosite number control (number of phosphosites passing QC), Sample quality control (coefficient of variance, CV, for technical and biological replicates). All criteria had to be met to receive a green QC flag.

QC levels were determined as follows:

Level	Flag1	Signal Strength (AU)	QC-passed phosphosites (PTK) of 195	Technical CV	Biological CV
Good		0–4,000	> 123	< 20%	< 30%
Fair		0–2,000	78–123	20−30%	30−40%
Poor		0–1,000	< 78	> 30%	> 40%

Phosphosite QC Selection: For the PTK assay, a kinetic readout was performed with images acquired every 5 min. Phosphorylation kinetics were analyzed by plotting S100 values against time, and only phosphosites exhibiting kinetics (signal increase over time) in at least 25% of the arrays (kinetic fraction > 0.25) were included in downstream analysis. Normalization was performed using ComBat correction, which applies empirical Bayes adjustments to reduce systematic bias between PamStation runs and PamChips.

Proteomic sample preparation

To analyze the proteomic profiles of the electroporated with a scrambled sgRNA (negative control) marked as wild type (WT) and XPC KO N/TERT-2G keratinocytes under basal conditions and after UVB irradiation, four distinct samples were prepared in quadruplicates (N = 4): wild-type non-irradiated, wild-type irradiated, XPC KO non-irradiated, and XPC KO irradiated. Cells were cultured in 6-well plates until reaching 80% confluence, then either irradiated or left untreated and incubated for 24 h at 37 °C. Following treatment, cells were trypsinized, and lysed with LYSE-NHS buffer (PreOmics, Germany). Samples were sonicated using a Bioruptor in a 4 °C water bath for 10 min, following 30-second cycles with 30 s on and 30 s off at maximal power. The sample mixture was then centrifuged at 16,000 rpm for 15 min at 4 °C. The supernatant containing the solubilized proteins was recovered. Protein quantification was performed using either the BCA (bicinchoninic acid) or Smith assay, both of which offer high sensitivity and minimal interference from other substances. Prior to analysis, 10 µg of protein per sample was loaded onto an SDS-PAGE gel to confirm the protein dosage, followed by Coomassie Brilliant Blue staining as per the manufacturer’s instructions (Life Technologies, California, USA). Samples were then prepared using the iST-NHS kit (Preomics), following manufacturer’s instructions. Peptides resulting from LysC/trypsin digestion were labelled using TMTpro™ 16plex Label Reagent Set (ThermoFisher Scientific) before mixing equivalent amounts for further processing. The peptide mix was then fractionated using the Pierce High pH Reversed-Phase Peptide Fractionation Kit (ThermoFisher Scientific).

LC–MS/MS and proteomic data analysis

Eight high-pH reversed-phase fractions were subjected to online nano-liquid chromatography coupled to tandem mass spectrometry (nLC–MS/MS) using an Ultimate 3000 RSLCnano system interfaced with an Orbitrap Exploris mass spectrometer (Thermo Fisher Scientific). Peptides were first trapped on a PepMap C18 precolumn (300 μm × 5 mm, Thermo Scientific) and subsequently separated on a 75 μm × 250 mm Aurora Generation 2 C18 analytical column (1.6 μm, IonOpticks) employing a 120-min acetonitrile gradient. Data acquisition was performed with Xcalibur software (Thermo Fisher Scientific).

Peptide and protein identification, as well as quantification, were carried out using MaxQuant (v1.6.17.0; PMID: 19029910)³⁵ against the UniProt Homo sapiens reference proteome (downloaded 20220322) supplemented with the MaxQuant contaminant database. Trypsin was specified as the digestion enzyme, permitting up to two missed cleavages. Fixed and variable modifications included carbamidomethylation (C), N-terminal acetylation, and methionine oxidation, respectively. The minimum peptide length was set to six amino acids, and at least one razor peptide was required for protein quantification. False discovery rates (FDRs) were controlled at 1% at both peptide and protein levels using a target-decoy approach.

Quantitative proteomic data were analyzed in ProStaR (PMID: 27605098)³⁶. Proteins identified as contaminants, in reverse databases, or only by site, as well as those quantified in fewer than four replicates per condition, were excluded. Reporter abundance values were log₂-transformed and normalized using variance-stabilizing normalization (VSN). Differential expression between conditions was assessed using limma, applying thresholds of |log₂(fold-change)| > 0.3 and p < 0.01, corresponding to an estimated FDR below 1.5% (Benjamini–Hochberg correction).

Bioinformatic analysis

All data were arranged based on the significance value and fold change so that XPC KO irradiated dysregulated proteins were used as a reference compared to the three other wild-type non-irradiated, wild-type irradiated, and XPC KO non-irradiated to decipher dysregulated proteins either upregulated, downregulated, or unique for XPC KO irradiated cells so that we can unravel their proteomic signature. Several bioinformatics tools were utilized to process the datasets, including StringApp (1.7.0), which makes the visualization and analysis of biological networks better, including all interacting biomolecules. The display of protein-protein interactions can be critically analyzed, and subnetworks can be created for specific interactions and correlations. The String enrichment feature conditions the enrichment analysis and annotates data such as Gene Ontology (GO). Associating the proteome network to different biological functions, such as biological processes, KEGG pathways, Reactome pathways, diseases etc., is instrumental in deciphering the network and bodily functions. FunRich (3.1.3) tool was employed to produce respective heatmaps.

Immunoblotting

Total protein was extracted by adding 100 µL of RIPA lysis buffer (Sigma-Aldrich, Missouri, USA) supplemented with phosphatase and protease inhibitor cocktails. Samples were incubated on ice for 30 min, transferred to 1.5 mL microcentrifuge tubes, and centrifuged at 16,000 rpm for 15 min at 4 °C. Protein concentration was determined using the BCA Protein Assay Kit (Life Technologies, California, USA). Western blotting was performed as previously described. Equal amounts of protein were separated by SDS–PAGE (Life Technologies) and transferred onto nitrocellulose membranes using the iBlot transfer system (Life Technologies). Membranes were blocked with 5% non-fat dry milk or bovine serum albumin (BSA) and incubated overnight at 4 °C with primary antibodies against XPC (Santa Cruz, 1:500), IFIT1 (Abcam, 1:500), IFIT3 (Santa Cruz, 1:500), IRF9 (Santa Cruz, 1:1000), ISG15 (Santa Cruz, 1:500), MX1 (Abcam, 1:500), MX2 (Abcam, 1:500), pSTAT1 (Tyr701; Cell Signaling Technology, 1:1000), pSTAT2 (Tyr690; Cell Signaling Technology, 1:1000), and pSTAT3 (Tyr705; Cell Signaling Technology, 1:1000). Following washes, membranes were incubated for 1 h at room temperature with HRP-conjugated secondary antibodies (mouse or rabbit; 1:5000–1:10000). Protein bands were visualized using Western Lightning ECL Pro reagent (PerkinElmer) and imaged with a Bio-Rad ChemiDoc™ XRS system. Band intensities were quantified using Image Lab™ software, with GAPDH serving as the loading control. All experiments were performed in triplicate (N = 3).

Statistical analysis

Statistical analyses, data normalization, and normality assessments were performed using GraphPad Prism v8 to determine the appropriate statistical test (non-parametric) for each experimental dataset.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

A.N. performed all the experiments and wrote the manuscript. E.S. and W.R. supervised the project. F.K. and H.R. revised the manuscript. Y.C. and L.B. performed M.S.-based quantitative proteomics. M.S., S.B.V., M.F.K. aided in bioinformatic analysis. All authors edited, read, and approved the manuscript.

Funding

AN is supported by a fund from the doctorate school (EDISCE) at University Grenoble Alpes. WR’s contribution was funded by ANR grant PG2HEAL (ANR-18-CE17-0017) and supported by the French National Research Agency in the framework of the “Investissements d’avenir” program (ANR-15-IDEX-02). HR’s and WR’s contribution were funded by ANR grant ANR-23-CE14-0045-01. The proteomic experiments were partially supported by Agence Nationale de la Recherche under projects ProFI (Proteomics French Infrastructure, ANR-10-INBS-08) and GRAL, a program from the Chemistry Biology Health (CBH) Graduate School of University Grenoble Alpes (ANR-17-EURE-0003).

Data availability

All data generated or analysed during this study are included in this published article (and its Supplementary Information files). The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (PMID: 39494541) partner repository with the dataset identifier PXD069309.

Declarations

Competing interests

The authors declare no competing interests.