We utilized data from The Cancer Genome Atlas (TCGA) project, which provides comprehensive molecular profiles for over 11,000 tumors across 33 different cancer types. Patient samples were selected based on the availability of both clinical and genomic data (5). Inclusion criteria included having complete information on ERCC gene expression, clinical outcomes, and relevant clinicopathological features. Samples with missing data or poor-quality genomic profiles were excluded to ensure the reliability of the analysis. These findings have the potential to provide valuable insights for precision oncology (15). Furthermore, a multitude of autonomous researchers have utilized The Cancer Genome Atlas (TCGA) as a valuable asset to bolster their own investigations and facilitate the analysis of molecular testing conducted on individual patients within a clinical context (16). Important data such as transcriptional expression and clinical relevance of 33 pan-cancer tumors were obtained through public data from the UCSC Xena database (17).

Initially, we eliminated any data that was not related to transcription and kept only the transcriptional data collected from 33 tumors that span across different types of cancer. Our objective was to generate a box plot that would facilitate a comparison of the expression levels of seven distinct ERCC genes (ERCC1-6 and ERCC8). Following that, a statistical analysis was conducted on the results of comparing gene expression in normal and tumor cells across 18 different forms of cancer. This study was confined to cases where the sample size for the normal group was greater than five. In the final phase of this research endeavor, the investigation focused on evaluating the extent of differential expression exhibited by the ERCC genes across 18 distinct types of malignancies. Additionally, the “corrplot” R package was utilized to conduct a more in-depth analysis of the correlation among ERCC genes.

The categorization was achieved by comparing their expression levels to the median level of expression of the ERCC genes. The differential analysis was also used to find out how the first treatment affected each cancer and how the mean levels of ERCC expression changed during different stages of the diseases. It was thought that there was statistical significance when the significance level was P < 0.05.

The cBioPortal for cancer genomics, accessible at http://cbioportal.org, is an online tool that allows users to explore, visualize, and analyze complex cancer genomics data. The portal translates data from molecular profiles of cancer tissues and cell lines into a series of events such as genes that can be easily interpreted. Personalized and customized data storage enables researchers to conveniently explore genetic variation and establish links to clinical outcomes when the underlying data is accessible (18). The portal offers many important aids to scientific research, most notably its ability to provide visualizations of gene-level data from multiple platforms. The user-friendly web interface of the portal enables researchers and physicians to access complicated cancer genomics profiles without the need for bioinformatics skills, hence facilitating biological discoveries (19). The dataset used in this investigation was collected from the ICGC/TCGA Pan-Cancer Analysis of Whole Genomes Consortium 2020. It comprised 2683 cases and was sourced via cBioPortal.

Our study aimed to examine the association between the expression of ERCC genes and immunological subtypes across various cancer types. The investigation also encompassed an examination of the immunological and stromal scores associated with each tumor sample. Subsequently, correlation analyses were conducted to correlate the associations between ERCC and immune/stromal scores.

The Spearman correlation test was utilized to ascertain the relationships between TMB/MSI and ERCC expression in pan-cancer. Furthermore, this test was employed to assess the correlations between the levels of immunological checkpoints expression and ERCC expression. We then created heatmaps with strong visual effects to better demonstrate the results of the analysis.

With TIMER 2.0, we performed a multidimensional systematic evaluation of different immune cells. The website can be reached at http://timer.comp-genomics.org/. We then analyzed whether there was a relationship between ERCC and different immune cell.

We obtained data on drug sensitivity and other relevant data from the CellMiner database. Afterwards, we then systematically analyzed the relationship between ERCC and drug sensitivity. Ultimately, a correlation was established between the manifestation of ERCC and the responsiveness to pharmaceutical agents.

GSEA was performed to determine the biological role and probable signaling mechanism of ERCC genes in the context of bladder cancer. This objective was achieved by utilizing the KEGG and Hallmark gene sets.

The statistical significance between different subgroups was assessed using the log-rank test, and the comparison of survival rates between groups with high and low expression levels was conducted by Kaplan-Meier curve analysis. To adjust for multiple testing, we applied the Benjamini-Hochberg procedure. Additionally, we utilized Cox proportional hazards models with multivariate adjustments for potential confounders, including age, gender, and cancer stage, to verify the prognostic significance of ERCC gene expression. All analyses were performed using R software, with an adjusted p-value < 0.05 considered statistically significant.

Figure 1 presents a comprehensive flow chart outlining the various stages and processes involved in this study. Significant variability in ERCC gene expression was observed across different cancer types. Initially, we conducted an examination of mRNA expression levels of seven ERCC genes (ERCC1-6 and ERCC8) in various cancer tissues. This investigation was carried out individually for each gene, and the results were compiled using the TCGA database ( Supplementary Figure S1 ). The expression levels of each ERCC gene were integrated across different types of cancers. To validate our findings, we utilized Cox proportional hazards models with multivariate adjustments, accounting for potential confounders such as age, gender, and cancer stage. This approach confirmed the robustness of the observed associations between ERCC expression and clinical outcomes. As depicted in Figure 2A , it was observed that ERCC1 exhibited higher expression in cancer compared to the other genes. Conversely, ERCC4, ERCC6, and ERCC8 demonstrated relatively low expression levels. Following that, a significant amount of variation in ERCC genes was detected among 18 different types of tumors. The inquiry subsequently analyzed the contrasting levels of expression of ERCC1-6 and ERCC8 in both cancerous and comparable healthy tissues. A total of 33 cancer samples were included in the analysis, with the exclusion of cancers for which the available normal tissue samples were fewer than 5. According to the data presented in Figure 2B , the expression levels of ERCC1 and 3 were found to be significantly elevated in CHOL samples compared to the corresponding normal tissues. The levels of ERCC1–6 and ERCC8 expression in KICH were significantly reduced in comparison to the expression observed in corresponding normal tissues. As illustrated in Figure 2C , a notable positive correlation (correlation coefficient = 0.47) is observed between ERCC1 and ERCC2, while a striking negative correlation (correlation coefficient = -0.2) is evident between ERCC1 and ERCC6. The results of this study indicate that there are inherent variations in the expression of ERCC genes, either among different ERCC genes themselves or among various types of cancer. Consequently, it is imperative to conduct comprehensive investigations on each of the ERCC genes, as they have the potential to serve as either oncogenes or anti-oncogenes in various types of cancer. Supplementary Figure S2 illustrates that several ERCC genes displayed higher levels of expression in distinct forms of cancer. For instance, ERCC1 demonstrated increased expression in cancerous tissues across all examined cancer types, surpassing its expression levels in normal tissues. ERCC2 expression levels were significantly higher in bladder cancer than in its corresponding normal tissues. Certain ERCC genes exhibited relatively low expression levels in specific tumor types. For instance, ERCC4 demonstrated low expression in KIRC, KIRP, PRAD, THCA, and UCEC. The levels of expression of ERCC genes exhibited significant variability within and between different types of cancer. Significantly, there was a lack of consistent pattern of expression seen among the several ERCC genes. For instance, the gene ERCC3 was found to be downregulated only in KICH.

Univariate Cox regression and Kaplan-Meier survival analyses revealed significant associations between ERCC expression and overall survival (OS), disease-free survival (DFS), disease-specific survival (DSS), and progression-free survival (PFS) across multiple cancer types. As shown in Figure 3A , there was a significant correlation between overall survival (OS) time and ERCC expression in multiple cancers. The results of the univariate Cox regression analysis demonstrated comparable prognostic outcomes when applying the same methodology to assess DFS, DSS, and PFS in a cohort of 33 tumors from TCGA dataset ( Figures 3B-D ). The relationship between each gene of ERCC and each cancer type has been further elucidated, as evidenced by the findings presented in Supplementary Table S1 . For instance, a correlation was observed between decreased expression of ERCC1 and unfavorable overall survival (OS) in ten types of cancers, namely BRCA, CESC, CHOL, ESCA, GBM, HNSC, LUSC, PAAD, PCPG, and PRAD. Conversely, elevated expression of ERCC1 was associated with poor OS in ten different cancers, specifically ACC, KICH, KIRC, LAML, LGG, LIHC, LUAD, MESO, SARC, and UCS. Subsequently, we examined the correlation between ERCC expressions and different stages of cancer. As shown in Supplementary Figure S3 , a clear negative correlation was observed between the mRNA expression of ERCC4 and the stage of KIRC. This indicates that as the stage of KIRC decreases, there is an increase in the mRNA expression level of ERCC4. According to Supplementary Figure S4 , following the initial treatment, the levels of ERCC 1, 2, 3, and 6 expressions were observed to be elevated in patients with adrenocortical carcinoma (ACC) who experienced progressive disease/stable disease (PD/SD) compared to those who achieved complete response/partial response (CR/PR). Furthermore, it was observed that patients with LGG exhibited elevated levels of ERCC4 expression in cases of progressive disease or stable disease, compared to cases of complete response or partial response. Conversely, the expression of ERCC5 was found to be lower in cases of progressive disease or stable disease, as opposed to cases of complete response or partial response.

Analysis of targeted sequencing data revealed a high frequency of genetic changes in ERCC genes, with amplifications being the predominant form. Based on the analysis of targeted sequencing data from a cohort of 2683 cancer patients, it was observed that the ERCC gene family exhibited alterations in 16.44% of the cases. Specifically, the ERCC1 gene showed alterations in 5% of the cases, ERCC2 in 5%, ERCC3 in 1.9%, ERCC4 in 3%, ERCC5 in 6%, ERCC6 in 2.6%, and ERCC8 in 4%. Additionally, mutations were detected in 2.68% of the cases (72 cases), amplifications in 11% (295 cases), deep deletions in 1.98% (53 cases), and multiple alterations in 0.78% (21 cases). These findings indicate a high prevalence of mutations within the ERCC gene family ( Figures 4A, B ). As depicted in Figure 4C , amplification emerges as the predominant form of genomic alteration within the ERCC gene family, exhibiting detectability across a diverse range of cancer types including lung cancer, endometrial cancer, ovarian cancer, bladder cancer, among others. Patients with alterations in ERCCs in pan-cancer cases may experience reduced therapeutic benefits when stratified based on these alterations. Furthermore, their median overall survival (OS) was significantly lower compared to patients without alterations in ERCC family genes (median OS: 39.05 months versus 55.43 months, P = 0.0227) ( Figure 4D ). Supplementary Figure S5 contains additional information regarding the mutations of each member of the ERCC family. Based on the findings presented in Figure 5 , it can be observed that patients with pan-cancer exhibit a higher prevalence of amplification and gain. Additionally, it is evident that the majority of ERCCs expressions show an upward trend as the copy number of CNV increases. Additionally, it was noted that the proportion of non-mutated regions was greater than that of missense mutations ( Supplementary Figure S6 ). The findings of the analyses suggest that copy number variations (CNVs) in pan-cancer are predominantly associated with the upregulation of ERCC genes.

As shown in Figure 6A , we used spearman correlation heatmaps to investigate the correlation between immune cell infiltration and ERCC expression in tumors. Research on the tumor immune microenvironment has garnered significant attention in recent years due to its potential to enhance our understanding of carcinogenesis, development, diagnosis, prevention, and predictive survival in patients. Upon analyzing the immune-related scores of each tumor sample, we discovered a notable relationship between ERCC differential expression and immune-related scores. To provide precise details, it was observed that ERCC2 exhibited a negative correlation with a majority of the 33 types of cancers ( Figures 6B, C ). This implies that an elevated expression of ERCC2 was associated with a reduced presence of immune cells within the tumors of said patients. Finally, in this study, RNA sequencing (RNAss) and DNA sequencing (DNAss) techniques were employed to investigate the correlation between the ERCC gene family and tumor stemness indices across various types of cancer. Various cancer types have been associated with the involvement of different members of the ERCC family in RNA and DNA processing. As shown in Figure 6D , there was a positive correlation between the expression of ERCC2 and most carcinomas, with only a mild negative correlation observed with OV. Similarly, Figure 6E demonstrated that at the DNA stem cell index level, the expression of ERCC2, 5, and 6 exhibited a significant positive correlation with OV, while showing a significant negative correlation with ERCC8.

Tumor immunotherapy has gained significant attention in recent years since it is closely associated with the use of immunosurveillance to predict predictive survival in cancer patients (20). We identified a set of three immune checkpoints, in order to investigate the relationship between ERCC expression and immune checkpoints. Our analysis revealed a relationship between ERCC differential expression and immune checkpoints ( Figure 7A ; Supplementary Figure S7 ). Tumor microenvironment (TMB) and microsatellite instability (MSI) are widely recognized as pivotal biomarkers in the context of tumor progression and carcinogenesis. TMB, an acronym for tumor mutational burden, refers to the density of nonsynonymous mutations within the protein coding region. This metric is closely associated with both tumor prognosis and the response to immunotherapy. We found that ERCC1 expression was significantly associated with TMB in ACC, GBM, KICH, STAD, SARC, LIHC, and LGG ( Figure 7B ; Supplementary Figure S8 ). Microsatellite instability (MSI) is a genetic condition characterized by the presence of insertion or deletion mutations that result in variable lengths of the microsatellite (MS) sequence during DNA replication. Functional problems with the mismatch repair system (MMRs) are a common cause of this phenomenon. It was observed that ERCC1 expression exhibited a significant correlation with MSI in various cancer types, including THCA, STAD, SARC, MESO, LIHC, KIRC, HNSC, GBM, DLBC, and BRCA. This finding is illustrated in Figure 7C and Supplementary Figure S9 .

To investigate the relationship between drug sensitivity and ERCC differential expression, we conducted comprehensive studies. The significance of ERCC family genes in cancer therapy resistance has gained increasing attention in the public sphere. The results revealed a significant correlation between ERCC and various medication sensitivities. According to the findings depicted in Figure 8 (p < 0.05), a positive correlation was observed between ERCC5 expression and the therapeutic effectiveness of PX-316, Chelerythrine, Palbociclib, Nelarabine, and Imexon. This implies that patients exhibiting higher levels of ERCC5 expression demonstrated improved treatment outcomes when administered these medications. Furthermore, a negative correlation was obtained between ERCC6 and the expression of Pralatrexate, Methotrexate, RH1, Belinostat, and Thiotepa. This suggests that patients with elevated ERCC6 expression should exercise caution when considering treatment with these drugs, as higher levels of ERCC6 expression are associated with reduced efficacy of these medications.

Several clinical trials have demonstrated that immune checkpoint inhibitors exhibit durable responses and have acceptable safety profiles over the long term. Nevertheless, it is worth noting that a significant proportion, ranging from 70% to 80%, of patients may continue to exhibit a lack of response to immune checkpoint inhibition. Therefore, further research is necessary to investigate the potential synergistic effects that may arise from the integration of immune checkpoint inhibitor therapy with other therapeutic modalities. This exploration aims to enhance the effectiveness of immunotherapies (21). In order to determine the influence of ERCC expression on the success of immunotherapy in different groups, we conducted a study to examine the relationship between ERCC gene expression and the effectiveness of immunotherapy in bladder cancer (GSE111636 and IMvigor), melanoma (GSE78220), and renal cell carcinoma (GSE67501) cohorts. A favorable connection was consistently seen between the expression levels of ERCC4 and ERCC5 and the objective responses to anti-PD-1/PD-L1 treatment in both the Imvigor210 and GSE78220 cohorts ( Figure 9 ). In contrast, it was observed that ERCC8 exhibited a positive correlation with resistance to cancer immunotherapy (specifically, anti-PD-1/PD-L1) in the datasets GSE111636 and GSE67501. In summary, the findings of this study suggest that the expression of ERCC genes may serve as a valuable predictor for the response to anti-PD-1/PD-L1 immunotherapy.

GSEA was used to study ERCC-related molecular pathways. Using this as a basis, we identified Hallmark and KEGG pathways that appear enriched in each ERCC. Based on the results obtained from statistical analysis using P value and FDR adjusted p-value, it was observed that a diverse range of tumor-related and immune-related Hallmark and KEGG pathways exhibited differential enrichment across various human cancers, with varying levels of ERCC expression ( Supplementary Figure S10 ). The presence of inconsistent results across various tumor types implies that ERCC genes may be involved in tumor progression through diverse mechanisms.

The work aimed to investigate the relationship between ERCC genes and tumor immunity, specifically examining the potential connection between immune cells and ERCC genes. Previous studies have demonstrated that human cancers exhibit six distinct immunological subtypes (C1-C6), each of which has the potential to either facilitate or impede tumor progression. The stages of wound healing are represented by C1-C6, with each stage corresponding to specific immunological processes. C1 signifies the dominance of interferon-r in the immune response, while C2 represents the inflammatory phase. C3 is characterized by lymphocyte failure, followed by C4 which indicates immune quiescence. In conclusion, the presence of C5 is linked to the prevalence of transforming growth factor-beta (TGF-) during the process of wound healing. Prior studies have demonstrated that individuals exhibiting immunological subtypes C3 and C5 exhibited significantly higher rates of survival in comparison to those with alternative subtypes, namely C4 and C6, which displayed the lowest rates of survival (22). The expression levels of the ERCC gene varied significantly across C1–C6, as depicted in Supplementary Figure S13A . The above findings suggest that ERCC is able to recognize several immune subtypes in cancer, thus aiding in cancer diagnosis. In order to obtain a deeper understanding, we examined the correlation between ERCC expressions and the extent of immune cell infiltration in pan-cancer. Further examination could reveal that ERCC1, 2, 3, 4, and 8 expressions were significantly higher in the C4 and 5 subtype compared to the other subtypes, implying a potential association between these genes and the promotion of cancer. In addition, we conducted a thorough examination of the relationship between ERCC and distinct clinicopathological characteristics in BLCA patients. The characteristics encompass tumor grade, and tumor subtype, among other factors ( Supplementary Figures S13B, C ). Our investigation revealed that several factors exert a substantial influence on the ERCC expression.

As shown in Figure 10 , there is a significant positive relationship between the expression of ERCC2, 3, 4, 6, and 8 and the BLCA stemness indices in BLCA patients, as evaluated by RNAss. Conversely, ERCC5 exhibits a negative correlation with RNAss (p < 0.05), while the association between ERCC1 and RNAss did not reach statistical significance. In our study, we found that ERCC 2 and 6 were significantly positively correlated with DNAss (p < 0.05). Nevertheless, the analysis did not reveal any statistically significant association between other ERCCs and DNAss (p > 0.05). According to the data presented in Figure 11 , there is a positive correlation between the expression of ERCC1, 2, 3, and 8 and the infiltration of CD8+T cells. Conversely, ERCC4, 5, and 6 exhibit a negative correlation with the infiltration of CD8+T cells. The findings of our investigation suggest that ERCCs may have the capacity to regulate the infiltration of immune cells in pan-cancer, which could have implications for the overall survival outcomes of individuals.

To further understand how ERCC affects immunotherapy, we conducted an investigation into the association between ERCC copy numbers and the abundance of commonly observed immune cells. The presence of ERCC3 and ERCC4 deletions, along with the amplification of ERCC1, ERCC2, ERCC5, and ERCC6 at the arm level, as well as the high amplification of ERCC8, were found to be significantly correlated with a notable decrease in the abundance of CD8+ T cells (refer to Figure 12A ). As illustrated in Figure 12B , the mutant group demonstrates a higher cellular volume in several immune cells in comparison to the non-mutant group. In this instance, it has been observed that the presence of a mutation in the ERCC1 gene leads to an increased quantity of resting NK cells as determined by CIBERSORT analysis. Conversely, non-mutant ERCC1 exhibits a higher level of NK cells as determined by EPIC analysis in comparison to mutant ERCC1.

We analyzed the protein expression levels of ERCC1, 2, 3, 4, and 5 using human protein atlas (HPA). As shown in Figures 13A-E , protein expression of ERCC1, 2, and 5 was up-regulated in bladder cancer tissues and EGR1 was down-regulated in cancer tissues compared to normal bladder samples. Subsequently, mRNA expression levels of ERCC2 and 5 were examined in bladder cancer tissues and normal bladder tissues. As shown in Figure 13F , the qRT-PCR results for these genes were similar to the protein expression results described above.

To investigate the role of ERCC2 in cell proliferation in bladder cancer cells, three siRNAs (si-ERCC2-1, si-ERCC2-2, si-ERCC2-3) were designed to silence ERCC2 expression in T24 cells. As shown in Figure 14A , all the above three siRNAs could effectively knockdown ERCC2 expression in T24 cells. To explore the effect of ERCC2 on bladder cancer T24 cells, we performed CCK-8 assay, colony formation assay, and three-dimensional multicellular spheres on the treated T24 cells. The CCK-8 results showed that the proliferative capacity of the experimental group in which ERCC2 was knocked down was significantly lower than that of the T24 cells in the NC group at 24, 48, 72, and 96 hour (p<0.05, Figure 14B ). As shown in Figure 14C , the results of colony formation experiments were similar to those of CCK8 experiments, and the proliferation capacity of cells in the NC group was significantly greater than that in the experimental group. Then three-dimensional (3D) multicellular spheres were introduced to continue to explore the effect of ERCC2 on the proliferation ability of T24 cells. As shown in Figure 14D , after prolonged incubation, the 3D multicellular spheres in the control group still maintained rapid growth, whereas the spheres in the experimental group did not show any significant growth, and the proliferative ability was weakened. In addition, live/dead analysis by applying calcein-AM/PI co-staining revealed the effectiveness of ERCC2’s loss of function in inhibiting the proliferation of T24 cells ( Figure 14E ).

To further explore the effects of ERCC2 on the migration and invasion of T24 cells, we performed wound healing assay and Transwell assay on T24 cells that were knocked down ERCC2. As shown in Figure 15A , the wounds in the control group healed significantly with the extension of time, while the experimental group did not show any significant wound healing, which indicated that the migratory ability of T24 cells after knocking down ERCC2 was significantly inhibited. Then the results of Transwell assay also confirmed the above conclusion that the migration and invasion abilities of T24 cells were significantly attenuated after knocking down ERCC2 ( Figure 15B ).The above results indicated that ERCC2 could effectively inhibit the migration and invasion of bladder cancer T24 cells.