The Potential of ESCO2 as a Prognostic and Immunotherapeutic Marker of Pan-Cancer and Its Role in Anti-PD-1 Treatment of Bladder Cancer

Bladder cancer (BLCA) has emerged as the most prevalent malignant tumor affecting the urinary system, representing approximately 50% of newly diagnosed tumors in this area [1]. This form of cancer can be categorized into two main groups: non-muscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC). Among individuals with MIBC, around half will encounter a recurrence within 2 years, and about 10% will already have distant metastasis upon initial diagnosis [2]. The standard treatment approach for localized MIBC typically involves cisplatin-based neoadjuvant chemotherapy followed by radical cystectomy [3]. Despite undergoing this surgical procedure, many patients continue to face the challenge of tumor recurrence and metastasis. Should the disease progress to the stage of distant metastasis, the 5-year survival rate drops significantly to a mere 8% [4]. In such critical scenarios, immunotherapy has emerged as a vital option for extending the lifespan of these individuals. One of the most remarkable advancements in current immunotherapy lies in the utilization of immune checkpoint inhibitors (ICIs), particularly in the treatment of BLCA. Due to this advancement in cancer treatment, the cancer death rate had fallen continuously over the past 20 years [5]. ICI treatment has made a significant contribution in cancer treatment, such as programmed cell death protein-1 (PD-1) [6] and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) [7]. Pembrolizumab, anti-PD-1, has notably received authorization for its application as a primary treatment choice for advanced BLCA patients [8]. Moreover, several clinical studies have highlighted the effectiveness of immunotherapy in treating individuals with advanced BLCA [9‒12]. However, not all patients with malignancies can benefit from immunotherapy [13]. It is urgent to discover new potential biomarkers to predict the response and prognosis of cancer patients receiving ICI, especially for BLCA.

The cell cycle is a rigorously regulated mechanism that facilitates the duplication of genetic material and the rapid growth of cells. An abnormal acceleration of cell growth caused by disruptions in the cell cycle regulation is a defining feature of cancerous growths [14]. The disruption of the cancer cell cycle promotes the uncontrolled growth of cells, driven by heightened mitotic signaling, failure of inhibitory checkpoints, or both. Research has shown that targeting proteins involved in the cell cycle can effectively curb the progression of tumors [15]. In recent years, the relationship between cell cycle and immunotherapy is being established. The cyclic GMP-AMP synthase-stimulator interferon gene (cGAS-STING) pathway is a promising therapeutic target in the prevention and treatment of a wide range of cancers [16]. Cell cycle-targeted therapy demonstrates a powerful synergy with immunotherapy through the activation of the cGAS-STING pathway, resulting in significant therapeutic advantages [17]. In regard to BLCA, scientists are currently investigating additional variables that could play a role in controlling the cell cycle of cancer cells, aiming to gain a deeper comprehension of how the cell cycle is regulated in BLCA [18].

Establishment of sister chromatid cohesion N-acetyltransferase 2 (ESCO2), a highly conserved cohesion acetyltransferase, is responsible for the establishment of sister chromatid cohesion and is involved in the acetylation of adhesins during the S phase of mitosis, which plays a crucial role in cell cycle [19]. ESCO2 inactivation could cause a disease named Roberts syndrome, which has attracted publication attention [20]. In recent years, several studies have substantiated the pivotal role of ESCO2 in the etiology and progression of various malignancies. ESCO2 knockdown in human gastric cancer cell lines in vitro significantly inhibited cell proliferation and induced apoptosis by regulating P53 [21]. The role of ESCO2 in targeting epithelial-mesenchymal transition in colorectal cancer had also been reported [22]. Besides, the ESCO2 expression was upregulated in aggressive melanoma [23] and breast cancer [24]. ESCO2 played a crucial role in controlling the advancement of low-grade glioma by influencing DNA replication by Liu et al. [25]. Furthermore, they observed a positive association between ESCO2 levels and the presence of different types of immune cells infiltrating the tumor microenvironment. It is crucial to conduct further investigations into the biological role of ESCO2 in various types of tumors, with a specific focus on the relationship between ESCO2 and immunotherapy.

To our knowledge, the potential relationship between ESCO2 and cancer immunotherapy at the pan-cancer level had not been reported yet. We evaluated the aberrant expression levels of ESCO2 across various cancer types and compared its expression between normal and tumor tissues. Additionally, we conducted analyses on clinical stage differences, prognosis, immune cell infiltration, and genomic alterations. Furthermore, we focused on BLCA and investigated the association between ESCO2 and immunotherapy using an openly accessible dataset. Moreover, we obtained imaging information and pathological sections from a BLCA patient treated with immunotherapy for validation. The most intriguing finding was that ESCO2 could serve as a predictive biomarker for immunotherapy response, highlighting its potential as a promising marker in this context. Our results suggest that ESCO2 is not only a prognostic biomarker for pan-cancer but also effectively predicts response to immunotherapy, thus providing valuable insights into the role of ESCO2 in cancer immunity.

We downloaded the standardized pan-cancer datasets TCGA Pan-Cancer (PANCAN, N = 10,535, G = 60,499) and TCGA TARGET GTEx (PANCAN, N = 19,131, G = 60,499) from the UCSC Xena database (https://xenabrowser.net/), and extracted the expression data of the ESCO2 (ENSG00000171320) in various samples. The samples screened included those from sources such as Solid Tissue Normal, Primary Solid Tumor, Primary Tumor, Normal Tissue, Primary Blood Derived Cancer Bone Marrow, and Primary Blood Derived Cancer Peripheral Blood. All gene expression data were standardized using the Transcripts Per Kilobase of Exon Model per Million Mapped Reads for all samples, and the log2(x + 1) transformation was applied to each expression value.

To further illustrate the differential expression of ESCO2 in normal and tumor tissues, we analyzed the expression data from the TCGA Pan-Cancer dataset and TCGA TARGET GTEx dataset, respectively. Cancer types with fewer than 3 samples were eliminated, resulting in the inclusion of expression data for 26 and 34 different cancer types in two datasets mentioned above. The abbreviations of cancers are represented in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000542188).

The clinical staging data corresponding to each sample were extracted from the TCGA Pan-Cancer dataset. To explore the differences in ESCO2 expression among samples from different cancers with various clinical stages, we performed a two-by-two analysis of the significance of differences using the Wilcoxon test and tested for differences in multiple samples using Kruskal-Wallis test. The variability of clinical features was conducted by the following three clinical staging metrics: tumor (T), lymph node (N), and metastasis (M).

In order to further investigate the prognostic role of ESCO2 in predicting malignant tumors, we chose the dataset with more prognostic information, TCGA TARGET GTEx, for our study. As for high-quality prognostic data, we obtained it from the article published by Liu et al. [26]. The sample screening criteria were set to Primary Blood Derived Cancer – Peripheral Blood (TCGA-LAML), Primary Tumor, TCGA-SKCM – Metastatic, Primary Blood Derived Cancer – Bone Marrow, Primary Solid Tumor, and Recurrent Blood Derived Cancer – Bone Marrow. Samples with follow-up times shorter than 30 days and the cancer with fewer than 10 samples were removed from data analysis process. Four types of prognosis features including overall survival (OS), disease-specific survival (DSS), disease-free interval (DFI), and progression-free interval (PFI) were assessed. Then, univariate Cox regression was used to assess the prognostic role of ESCO2 for a specific prognosis type in each cancer. ESCO2 expression levels were used to perform Kaplan-Meier curve analysis, with the cut-off chosen by the “surv-cutpoint” function of the “survminer” R package.

The immune cell infiltration levels of TCGA cancers were downloaded from the Tumor Immune Estimation Resource 2.0 (TIMER2) database (http://timer.cistrome.org/ [27]), which is a data source for quantifying immune cell infiltrations across distinct cancers. We analyzed the correlations between ESCO2 expression and 18 immune cell subsets including B cells, hematopoietic stem cells (HSCs), natural killer T (NKT) cells, macrophages, mast cells, monocytes, myeloid-derived suppressor cells (MDSCs), neutrophils, natural killer cells, T-cell follicular helper, γ/δ T cells, cancer-associated fibroblast, regulatory T cells, CD4+ T cells, CD8+ T cells, dendritic cells, endothelial cells (Endo), and eosinophil in pan-cancer. The R code for drawing the immune cell infiltration heatmap was provided by Tu et al. [28].

The web portal TISIDB database (http://cis.hku.hk/TISIDB/) [29] was used to analyze the correlation between ESCO2 expression with immunoinhibitors, immunostimulators, major histocompatibility complex molecule, chemokine, and receptor. Furthermore, we conducted the correlation of ESCO2 expression and CD274 (PD-L1) expression in BLCA. We calculated the TMB of each tumor using the “tmb” function of the R package “maftools” and integrated the TMB and gene expression data of the samples from TCGA Pan-Cancer dataset. As for MSI, we obtained the data from previous study [30] and combined the MSI score with paired samples. We analyzed the correlation between ESCO2 expression levels with TMB and MSI at the pan-cancer level to show the relationship between ESCO2 expression and tumor immunity.

To further explore the immune role of ESCO2 in the BLCA, the dataset GSE176307 [31] from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/), containing the RNA-seq and clinical data of immunotherapy-treated BLCA patients, was downloaded. The samples without complete RNA-seq and clinical data were eliminated. The “surv-cutpoint” method within the “survminer” R package was employed to certain the most suitable cut-off value. By utilizing the ESCO2 cut-off value, the samples were categorized into two subgroups based on their expression levels: high and low. Complete response (CR), partial response (PR), progressive disease (PD), and stable disease (SD) were provided by the GSE176307 dataset, with objective response rate (ORR) = [(CR+PR) cases/total cases] × 100%. The K-M survival curve was plotted to perform the progression-free survival probability. Besides, we also analyzed the correlation between TMB and ESCO2 expression level.

The fold change and adjusted p values of each gene were obtained from the GSE176307 dataset by performing differential expression analysis using the “limma” R package based on the ESCO2 expression. The top twenty-five genes exhibiting significant positive or negative correlations with ESCO2 were then extracted and depicted in a heatmap. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted using the “ClusterProfiler” R package.

The postoperative pathology specimens from a case of BLCA diagnosed in August 2023 and treated with immunotherapy were collected, including tumor tissues and tumor adjacent tissues. The images of enhanced computed tomography (CT) before and after immunotherapy treatment of the patient were collected, and the relevant informed consent forms had been signed. Additionally, tumor samples from 10 other BLCA cases at Qilu Hospital of Shandong University, which did not receive immunotherapy, were collected and analyzed, with the relevant informed consent forms having been signed.

Formaldehyde-fixed paraffin-embedded BLCA tumor tissues were cut into 5 µm sections. Following incubation at 60°C, the formaldehyde-fixed paraffin-embedded tissue slices underwent deparaffinization in xylene and rehydration in a series of graded ethanol solutions. Antigen retrieval was carried out using citrate buffer in a microwave for 15 min, followed by natural cooling to room temperature. Subsequently, the samples were treated with 3% H₂O₂ for 10 min and blocked with 3% bovine serum albumin for 30 min at 25°C to prevent nonspecific binding. After overnight incubation at 4°C with primary antibodies including ESCO2, CD3, CD4, CD8, and CD274 (PD-L1) from ABclonal, Wuhan, China, the samples were incubated with suitable secondary antibodies for 30 min. Detection of the reactions was achieved using a DAB detection kit (DAB+ ZSGB-BIO, China), with appropriate termination of staining and counterstaining with hematoxylin. The 10 BLCA patient specimens, without any prior immunotherapy, were stained with ESCO2 and CD274 using the aforementioned immunohistochemistry (IHC) methods. Additionally, the patient specimen who received immunotherapy was stained with ESCO2, CD3, CD4, CD8, and CD274.

The SW780 and MB49 BLCA cell lines were acquired from CellSource, China, and were limited to 15 passages for each experiment. Regular testing was conducted to check for any mycoplasma contamination. The cell culture media used for all experiments contained 10% fetal bovine serum (ExCell Bio, China) and 1% penicillin and streptomycin (KeyGen Biotech, China). Incubation of the cell lines was carried out at 37°C with 5% CO₂. The transfection reagent utilized in this research was jetPRIME (Polyplus-transfection, 101000046). To begin, combine 4 μg of plasmid or 20 nm of siRNA with 200 μL of transfection buffer solution in an EP tube, followed by the addition of 8 μL of transfection reagent. After thorough mixing, allow it to sit for 10 min. Then, transfer the prepared mixture to the culture dish and ensure thorough mixing. Monitor cell growth and morphological changes, and assess RNA and protein expression levels after either 24 or 48 h. The sequences were as follows: si-ESCO2 – 5′-CUC​UUA​GAC​CAG​GAU​UAU​Ctt-3′.

Cell viability was measured using the Cell Counting Kit-8 (CCK-8) according to the manufacturer’s instructions. Transwell chambers with an 8-μm pore size from Corning in Tewksbury, USA, were utilized to evaluate the migration and invasion capabilities of BLCA cells. In the Transwell migration assay, cells were seeded into the upper chambers of 24-well Transwell inserts at a density of 5 × 10⁴ cells per well and cultured in serum-free medium, while the lower chambers were filled with medium containing 10% FBS. After 24 h, the migrated cells were stained with a 1% crystal violet solution. The cells that migrated to the lower surface were then counted under a microscope and captured in photographs at ×200 magnification. For the invasion assay, the upper Transwell chambers were coated with 50 μL of Matrigel from Becton Dickinson in Franklin Lakes, NJ, for 4 h. After 24 h, the non-invaded cells in the upper inserts were eliminated, and the invasive cells were imaged and quantified using the same methodology as described earlier.

R software (Version 4.3.1, https://www.Rproject.org) was used for data analysis, and the “ggplot2” R package and the web portal SangerBox 3.0 (http://vip.sangerbox.com/) were applied to draw the plots [32]. Wilcoxon test was performed to compare continuous variables. Cox regression analysis and the Kaplan-Meier method were employed to assess the prognostic of ESCO2 expression in various cancers and GSE176307. Spearman correlation analysis was performed to assess the statistical relationships between ESCO2 and other factors. The χ² test was used to explore the difference of ORR in the high- and low-expression subgroup. In all statistical procedures, the p value <0.05 was considered statistically significant.

To examine the differential expression of ESCO2 between tumor and normal tissues, we conducted separate analyses of mRNA expression values in both the TCGA Pan-Cancer cohort and the TCGA TARGET GTEx cohort. Our findings indicated that ESCO2 expression was significantly upregulated in 18 different types of cancers, including BLCA, BRCA, CESC, CHOL, COAD, ESCA, GBM, HNSC, KIRC, KIRP, LGG, LIHC, LUAD, LUSC, STAD, STES, THCA, and UCEC (Fig. 1a). In the TCGA TARGET GTEx cohort, our findings showed consistently higher expression levels of ESCO2 in tumor tissues across various cancer types, with the exception of ACC, KICH, PCBG, and READ (Fig. 1b). These findings suggested that ESCO2 may play a key role in the diagnosis of cancer.

In the examination of the relevance of tumor stage, we investigated the variability in ESCO2 expression across different TNM stages. We have included only those analyses that resulted in statistically significant results according to the Kruskal-Wallis test (Fig. 2). For the T stage, significant differences were observed in 10 tumors, namely, ACC, BRCA, KIRC, KIRP, LIHC, LUAD, LUSC, PAAD, PRAD, and STAD. In terms of the N stage, differential expression was noted in BRCA, COAD, LUAD, KIRP, PRAD, and THCA. Regarding the M stage, elevated expression of ESCO2 was specifically identified in ACC-M1 and COAD-M0. This series of results indicated that ESCO2 could reflect, to some extent, the malignancy of tumors.

The survival association analysis was conducted for each cancer type, focusing on OS, DSS, DFI, and PFI. In terms of OS, high expression of ESCO2 was identified as a significant risk factor in the LGG, KIRP, ACC, KICH, MESO, LIHC, LAML, LUAD, TARGET-LAML, PAAD, TARGET-ALL-R, and TARGET-ALL cohorts. However, it exhibited a protective effect in THYM, COAD, and READ (Fig. 3a). To address the potential bias resulting from non-tumor deaths, we subsequently performed DSS analysis. The findings from the DSS analysis indicated that ESCO2 overexpression was linked to shorter DSS in KIRP, LGG, KICH, ACC, MESO, PAAD, LUAD, BLCA, and SKCM-P (Fig. 3b). Referring to associations between ESCO2 expression and DFI, high ESCO2 expression was found to be associated with poor DFI in KIRP, PAAD, THCA, LIHC, and PCBG (Fig. 3c). Furthermore, the PFI results demonstrated that high ESCO2 expression was correlated with poor prognosis in patients with KIRP, LGG, ACC, KICH, PAAD, MESO, LIHC, PCBG, BLCA, LUAD, BRCA, THCA, and PRAD (Fig. 3d). Additionally, a survival analysis based on four prognostic indicators was performed in the BLCA cohort. The survival curves indicated a strong association between upregulation of ESCO2 expression and reduced DSS (HR = 1.49, 95% CI: 1.02–2.16, p = 0.04; Fig. 3f) and PFI (HR = 1.56, 95% CI: 1.14–2.13, p = 4.7e−3; Fig. 3h). However, no statistically significant difference was observed for OS (HR = 1.29, 95% CI: 0.94–1.77, p = 0.11; Fig. 3e) and DFI (HR = 1.57, 95% CI: 0.75–3.29, p = 0.23; Fig. 3g) between the high and low subgroups. This series of results demonstrated the potential of ESCO2 as a predictive prognostic factor.

We investigated the correlations between ESCO2 expression and immune cell infiltrations based on the immune cell infiltration data from the TIMER2 database. The heatmap revealed the infiltration levels of B cells, HSC, NKT cells, macrophages, mast cells, monocytes, MDSC, neutrophils, natural killer cells, T-cell follicular helper, γ/δ T cells, cancer-associated fibroblast, regulatory T cells, CD4+ T cells, CD8+ T cells, dendritic cells, Endo, and eosinophil across multiple cancer types (Fig. 4). The results show that ESCO2 was positively correlated with the infiltration levels of B cells, CD4+ T cells, CD8+ T cells, MDSCs, and neutrophils across most TCGA cancers. On the other hand, ESCO2 expression was negatively correlated with the infiltration levels of HSCs, NKT cells, and Endo cells. Notably, high ESCO2 expression exhibited a positive correlation with Th2 cell infiltration in almost all TCGA cancers. These findings highlighted the significance of immune cell involvement in cancer immunotherapy. Additionally, our results suggested that ESCO2 may influence cancer development and prognosis by interacting with immune cell populations.

We investigated the correlations of ESCO2 expression with 24 immunoinhibitors (Fig. 5a), 45 immunostimulators (Fig. 5b), 21 major histocompatibility complex molecules (Fig. 5c), 41 chemokines (Fig. 5d), 18 chemokine receptors (Fig. 5e) using data from the TISIDB database. We observed a positive association between ESCO2 expression and the majority of immunoinhibitors in KIRC and THCA. Additionally, a robust positive correlation was identified between ESCO2 and CD274 in various TCGA cancers, with a particularly significant correlation in BLCA where the correlation coefficient was calculated to be r = 0.193 (Fig. 5f). To investigate the role of ESCO2 in predicting the effectiveness of ICIs, we conducted additional analyses to explore the correlations between ESCO2 expression and TMB and MSI. Positive correlations with TMB were identified in LGG, LUAD, COAD, BRCA, STES, SARC, STAD, PRAD, PAAD, OV, BLCA, ACC, and KICH, while negative correlations were observed in THCA and THYM (Fig. 5g). For the correlation between ESCO2 expression and MSI, positive correlations were identified in COAD, STES, SRAC, STAD, and MESO, and negative correlations were identified in HNSC, THCA, and DLBC (Fig. 5h). The results suggested that ESCO2 has potential as a predictor of ICI efficacy.

To investigate the association between ESCO2 and cancer immunotherapy for BLCA, we retrieved a BLCA immunotherapy cohort (GSE176307) from the GEO database. The patients with BLCA in this cohort received treatment with anti-PD-1 (nivolumab and pembrolizumab) or anti-PD-L1 (atezolizumab, avelumab, and durvalumab). A total of 88 available samples were included for analysis. We utilized the “surv-cutpoint” function from the R package “survminer” to determine the optimal cut-off value for ESCO2 expression. Based on this threshold, we divided the 88 samples into high-expression and low-expression groups, each consisting of 44 samples. Comparison of ORR between these two groups revealed a higher percentage of CR/PR in the high-expression subgroup; however, statistical significance was not reached (p = 0.097; Fig. 6a). In this dataset, we observed a positive correlation between ESCO2 expression and TMB (r = 0.24, p = 0.02; Fig. 6b). In contrast to the findings from the TCGA cohort analysis, our results demonstrated that high expression of ESCO2 was associated with a longer PFI (HR = 0.58, 95% CI: 0.37–0.91, p = 0.02; Fig. 6c). This suggested a dual role for ESCO2 as both an oncogene and an immune gene within this dataset. Furthermore, differential gene expression analysis identified a total of 3,196 upregulated genes and 1,681 downregulated genes which were visually represented using a volcano plot (Fig. 6d) based on the expression of ESCO2. Co-expression analysis revealed the top twenty-five genes exhibiting significant positive or negative correlations with ESCO2 (Fig. 6e).

Subsequently, GO and KEGG enrichment analysis was conducted to investigate the potential impact of ESCO2 on tumorigenesis by examining the enrichment pathways separately for up- and downregulated differentially expressed genes. KEGG pathway analysis revealed that ESCO2 is involved in various pathways, including DNA replication, homologous recombination, mismatch repair, and the cell cycle (Fig. 7a). Biological process enrichment analysis demonstrated that upregulated genes associated with ESCO2 were primarily implicated in DNA replication, chromosome segregation, and cell cycle processes (Fig. 7b). Furthermore, cellular component enrichment analysis indicated that these upregulated genes were enriched in condensed chromosome kinetochore structures (Fig. 7c). Molecular function enrichment analysis highlighted the significant role of ESCO2 in tumor pathogenesis through its association with single-stranded DNA-dependent ATPase activity (Fig. 7d). Bubble plots were used to individually present the results of GO and KEGG pathway enrichment analyses for ESCO2-related downregulated genes (Fig. 7e–h). The results of enrichment analysis indicated that ESCO2 regulated the occurrence and progression of tumors by participating in these pathways.

We created a chronological representation of the patient’s visit experience (Fig. 8a). In August 2023, the patient received a diagnosis of BLCA. The attached CT imaging revealed significant thickening of the inner wall of the patient’s bladder (Fig. 8a). Following this, the patient underwent three cycles of anti-PD-1 tumor immunotherapy using toripalimab. Subsequent CT imaging post-immunotherapy showed a remarkable reduction in tumor tissue size. In December 2023, a laparoscopic radical cystectomy was performed on the patient. Para-tumor tissue and tumor tissue were chosen for separate immunohistochemical staining analysis (Fig. 8b, c). Notably, the protein expression levels of the ESCO2 gene in tumor cells showed a clear increase compared to normal cells. Similar differences were noted in the expressions of CD3, CD4, CD8, and CD274. This finding suggests a co-expression relationship between ESCO2 and CD3, CD4, CD8, and CD274 proteins, indicating that ESCO2 may have potential as a predictor of immunotherapy response.

To demonstrate the correlation between ESCO2 and CD274 co-expression, we collected 10 BLCA samples that had not undergone immunotherapy. Subsequent immunohistochemical staining was conducted to examine the expression levels of ESCO2 and CD274. The IHC results indicated simultaneous alterations in the levels of ESCO2 and CD274 expression (Fig. 9a). Additionally, a Transwell assay was utilized to evaluate the migration and invasion capabilities of BLCA-derived cell lines, SW780 and MB49. The results showed a notable decrease in migratory and invasive cells following the knockdown of ESCO2 compared to the control group, while the migratory and invasive cells in the overexpression subgroups of both cell lines did exhibit an increase compared to the control group (Fig. 9b). The results of CCK-8 experiment showed that the proliferation of SW780 cell line inhibited by siRNA transfection to suppress ESCO2 expression was inhibited (Fig. 9c), while the proliferation promoted by plasmid transfection to overexpress ESCO2 showed the opposite result (Fig. 9d). Collectively, these findings provide compelling evidence supporting the enhancing impact of ESCO2 on BLCA cell invasion, migration, and proliferation.

ESCO2, a type of acetyltransferase, plays a crucial role in sister chromatid cohesion by modifying the SMC3 subunit of cohesion [33]. Genomic mutations in the sister chromatid cohesion pathway have been found to be significantly associated with the development of cancer [34]. Previous studies have shown differential expression of the ESCO2 gene in various malignancies compared to normal tissues, revealing its potential role in cancer. Zhu et al. [35] discovered that ESCO2 was upregulated in lung adenocarcinoma compared to adjacent normal lung tissues, and its overexpression promoted cell proliferation and metastasis. Additionally, Liu et al. [25] reported that ESCO2 was an independent risk factor associated with decreased OS in glioma patients. Our study conducted a pan-cancer analysis of ESCO2 and revealed significant overexpression of ESCO2 in BLCA tissues.

ICIs have become an important role in tumor immunotherapy [36]. However, immunotherapy may not be effective in all patients due to tumor heterogeneity. In order to make immunotherapy accessible to a larger number of tumor patients, researchers are investigating potential biomarkers. In this study, we have found that ESCO2 is a prognostic biomarker for pan-cancer, and it can effectively predict the response to immunotherapy, especially in BLCA.

First, we assessed the expression levels of ESCO2 across various cancer types by using data from the UCSC Xena database. Our results consistently showed a significant upregulation of ESCO2 expression in the majority of the analyzed cancers, indicating its potential role as an oncogene. Following this, we examined the differential expression patterns of ESCO2 among different clinical tumor stages. Importantly, our findings demonstrated a strong association between the mRNA levels of ESCO2 and clinicopathological stages in a range of malignancies.

When it comes to prognostic analysis, we investigated the association between ESCO2 and four prognostic characteristics in pan-cancer. Our findings from analyses of OS, DSS, DFI, and PFI consistently demonstrated a significant correlation between ESCO2 expression and cancer prognosis. Additionally, integrating the results of differential expression analysis revealed that high ESCO2 expression was associated with a poorer prognosis in multiple cancer types. Specifically focusing on BLCA, Kaplan-Meier survival curves further confirmed the detrimental impact of elevated ESCO2 expression on patient prognosis. Taken together, our findings suggest that ESCO2 holds promise as a potential prognostic biomarker for predicting cancer outcomes.

Immune infiltration is a significant feature of the TME, with the proportion of different immune cells, especially T cells, playing a crucial role in influencing antitumor immunity and immunotherapy outcomes. The fundamental principle of immunotherapy is to stimulate the body’s immune cells to target and attack cancer cells, thus aiming to achieve therapeutic goals [37]. Therefore, assessing the extent of immune cell infiltration in the TME is critical for predicting the effectiveness of tumor immunotherapy. In our study, we found that ESCO2 expression was positively associated with the levels of infiltration of CD4+ T cells, CD8+ T cells, MDSCs, and neutrophils in most cancers. The correlation analysis between ESCO2 and immune markers, particularly CD274, suggests the potential role of ESCO2 in immunotherapy across various types of cancer. The overexpression of CD274 has been shown to inhibit antitumor immune responses and promote tumor growth, proliferation, and survival in a variety of tumor types, including COAD, KIRC, BLCA, and HNSC [38]. Previous study showed that patients with high PD-L1 expression had a strong correlation with BLCA progression [39]. In a study by Inman et al. [40], the strong association between CD274 with high-grade tumors (OR = 2.4; p = 0.009) and tumor infiltration (OR = 5.5; p = 0.004) was evaluated. Above all, the positive correlations between ESCO2 and CD274 suggest the potential of ESCO2 as an immunotherapy biomarker.

TMB is highly correlated with the efficacy of PD-1/PD-L1 inhibitors [41], and MSI occurs as a result of functional defects in DNA mismatch repair in tumor tissues. The phenomenon of MSI accompanied by defective DNA mismatch repair is a clinically important tumor marker [42]. They are reliable biomarkers for prognosis in various cancers and predictors of many tumors’ immunotherapeutic effect [41, 43]. Our results showed a significant positive correlation between ESCO2 expression and TMB in 13 cancers, as well as with MSI in 5 cancers, including BLCA. Therefore, we hypothesized that higher ESCO2 expression may lead to improved survival benefits following immunotherapy.

GO and KEGG enrichment analyses revealed that ESCO2-related DEGs were mainly involved in DNA replication, homologous recombination, mismatch repair, and cell cycle. The presence and progression of cancer are significantly related to DNA mismatches and dysregulation of the cell cycle [44]. Dysfunction in DNA replication or chromosome segregation poses challenges in the initiation and development of cancer, while also presenting opportunities for cancer treatment [17]. Studies have shown that mutations in the homologous recombination repair gene are associated with the effectiveness of immunotherapy, and patients with these mutations have significantly improved prognosis following immunotherapy [45]. Besides, mismatch repair-deficient tumors have been reported to exhibit susceptibility to ICIs that target PD-1 and PD-L1 in BLCA [46]. With a better understanding of the factors that regulate the cell cycle in cancer cells, targeting and blocking these processes could potentially be a method for controlling cancer progression in the future.

As part of validation, we conducted the analysis on the immunotherapy cohort in BLCA. In the subgroup with high ESCO2 expression, we observed a higher percentage of CR or PR and longer PFI, indicating that ESCO2 serves as a robust prognostic biomarker in BLCA patients undergoing antitumor immunotherapy. In addition, immunohistochemical staining images from specimens of a BLCA patient who underwent immunotherapy demonstrated increased ESCO2 expression and co-expression between ESCO2 and CD274 in tumor tissues compared to adjacent non-tumor tissues. The co-expression of ESCO2 and CD274 was also observed in IHC-stained BLCA samples without immunotherapy. In vitro experiments, including CCK-8 and Transwell assay, provided compelling evidence supporting the enhancing impact of ESCO2 on BLCA cell invasion, migration, and proliferation.

Therefore, we propose that ESCO2 may serve as an effective biomarker for cancer immunotherapy, with significant potential for future applications in tumor treatment, particularly for BLCA. To be transparent, there are several limitations to our research. First, despite our best efforts to standardize data from various open-access databases, there may still be inherent systematic biases present. Second, the limited number of patients who underwent immunotherapy may compromise the overall reliability of the experiment. Moving forward, we plan to conduct animal experiments and delve into deeper research to investigate the correlation between ESCO2 and BLCA. We firmly believe that well-designed experiments are essential for the future clinical applications and mechanism investigations in this area.

The comprehensive pan-cancer analysis has revealed ESCO2’s potential as a prognostic biomarker for various cancers and its ability to predict responses to immunotherapy. Overexpression of ESCO2 is associated with prognosis, immune regulation, immune cell infiltration, TMB, and MSI in various types of cancer. The results of in vitro experiments indicate that ESCO2 regulates the proliferation, invasion, and migration of BLCA cells. Based on these findings, we suggest that targeting ESCO2 could be an effective therapeutic strategy for cancer treatment.

We sincerely acknowledge the contributions from the TCGA Pan-Cancer Project and SangerBox website.

All patients participating in this study have signed a written informed consent form. The studies involving human participants were reviewed and approved by the Ethics Committee of Qilu Hospital of Shandong University, Approval No. KYLL-2020(KS)-262.

The authors declare that the research was conducted in the absence of any commercial relationships that could be construed as a potential conflict of interest.

This work was supported by grants from the Department of Science and Technology of Jinan City (Grant No. 201805030).

J.L. and Y.F. designed the study and revised the manuscript. W.G. and S.Z. conducted the data collection, bioinformatic and statistical analysis, figure visualization, and manuscript writing. J.L. did the revision of the study. W.G. and K.Y. collected the patient pathological specimen. All coauthors have approved the final version of the manuscript.

Wei Guo and Shuo Zhao contributed equally to this work and share first authorship.

Publicly available datasets were analyzed in this study. These data can be found here: https://xenabrowser.net/datapages.