Abstract
Background/Aim: Clear cell renal cell carcinoma (ccRCC) is the most common subtype of kidney cancer with aggressive behavior and poor prognosis. Dysregulation of DNA replication and repair, including alterations in replication protein A1 (RPA1), replication factor C subunit 1 (RFC1), and DNA polymerase epsilon (POLE), may influence tumor biology and immune interactions. This study investigated the expression of these proteins in ccRCC and their associations with systemic inflammation, tumor immune microenvironment (TME), and prognosis.
Materials and Methods: Immunohistochemical expression of RPA1, RFC1, and POLE was evaluated in 52 ccRCC and adjacent normal tissues, with correlations to clinical data and preoperative blood parameters. Transcriptomic data from The Cancer Genome Atlas (TCGA) and immune deconvolution analyses (TIMER2.0, ConsensusTME) validated findings and explored associations with immune infiltration and survival.
Results: Tumor tissues showed increased RPA1 and decreased RFC1 expression, while POLE was unchanged. Elevated RPA1 correlated with reduced systemic inflammation, while low RFC1 correlated with larger tumor size. High POLE levels associated with lower preoperative platelet-to-lymphocyte ratio and an inverse trend with T stage. TCGA data confirmed these findings, showing that low RPA1 and RFC1 predicted poorer outcomes, while reduced POLE and RFC4 were linked to improved survival. TIMER2.0 analysis revealed that high RPA1 and RFC1 expression was linked to increased macrophage and neutrophil infiltration, whereas high POLE to CD4+ T-cell infiltration. Notably, RPA1, RFC1, and POLE expression correlated with immune-checkpoint molecules (including PD-L1, VISTA, and CTLA-4), suggesting implications for immunotherapy responsiveness. Immune TME composition, as estimated by ConsensusTME, influenced survival outcomes, with replication protein expression modulating prognostic relevance of immune subpopulations.
Conclusion: DNA replication proteins interact with systemic inflammation and the TME in ccRCC, supporting their role as biomarkers and potential therapeutic targets.
- Replication protein A1
- DNA polymerase epsilon catalytic subunit
- replication factor C subunit 1
- clear cell renal cell carcinoma
- tumor immune microenvironment
- DNA replication and repair
Introduction
Renal cell carcinoma (RCC) is the 13th most common malignancy worldwide, accounting for 2.2% of adult cancers (1). Clear cell RCC (ccRCC) is the most prevalent histological subtype, characterized by high morbidity and mortality (2). Although localized ccRCC can be predominantly cured with surgery, approximately one-quarter of patients present with metastatic disease at diagnosis (3), and despite advances in targeted therapies and immunotherapy with immune checkpoint inhibitors (ICIs), metastatic ccRCC remains largely incurable with short 5-year overall survival (OS) rates (4, 5). Unfortunately, a significant number of patients with advanced ccRCC fails to respond or will develop resistance to first-line therapies (6). Therefore, there is an urgent need to identify and validate biomarkers that provide predictive and prognostic information in the metastatic setting of ccRCC.
DNA replication and repair are complex, tightly controlled processes performed by multiple protein complexes that ensure stability of the genome (7). Dysregulation of these pathways can lead to genome instability, thus predisposing to cancer (7, 8). Three polymerases – α, δ and ɛ – in symphony with accessory replication proteins, carry out DNA replication (7). DNA polymerase ε (POLE) and δ (POLD) are four-subunit enzymes involved in DNA proofreading and replication on the leading and lagging strands, respectively, while polymerase α initiates DNA replication (9, 10). Replication protein A (RPA), a heterotrimeric protein, stabilizes single-stranded DNA, while a ring-shaped proliferating cell nuclear antigen (PCNA) acts as a sliding clamp for DNA polymerases that is loaded around DNA by replication factor C (RFC), a five-unit protein complex, altogether orchestrating a precise and efficient DNA replication process (11). Knowing that polymerases along with accessory proteins are crucial for accurate DNA replication and repair, their dysregulation may lead to genomic instability and uncontrolled cell proliferation characteristic of cancer. Although POLE and POLD1 mutations are relatively infrequent in kidney cancers (≈2.44%) (12), both germline and somatic alterations in these genes lead to defects in DNA repair. These defects cause high tumor mutational burden (TMB), increased neoantigen load, and elevated tumor-infiltrating lymphocytes in the tumor microenvironment (TME), all of which are associated with susceptibility to ICIs (13). Moreover, immunotherapy seems to be effective in POLE/POLD1-mutated tumors regardless of microsatellite status (14). While some studies suggest that elevated POLE and POLD1 expression may correlate with an immunosuppressive TME in ccRCC (15, 16), the prognostic and predictive value of the expression levels of replication accessory proteins remains to be fully elucidated. Our group previously showed that higher nuclear POLD1 expression in ccRCC correlated with longer survival, suggesting a favorable prognosis (17). However, another team linked increased POLD1 with poorer OS, highlighting the controversy over replication protein prognostic value (16). These conflicting results motivated us to investigate additional replication proteins – RPA1, RFC1, and POLE – to further elucidate their roles in ccRCC progression.
In addition to these molecular features, TME – the complex network of immune cells, fibroblasts, blood vessels, and signaling molecules surrounding tumor cells – plays a critical role in tumor progression and treatment response (18). Recent studies have highlighted that alterations in the immune cell composition within the TME, including variations in T cells, macrophage subtypes, and other immune effectors, are closely linked to ccRCC prognosis (19-21). Systemic inflammation, as reflected in preoperative blood parameters such as white blood cell count, hemoglobin, and platelet indices, also significantly influences tumor behavior and therapeutic outcomes. Emerging evidence indicates that systemic inflammatory markers measured in peripheral blood may be linked to the immune cell composition of the TME, raising the possibility that blood-based assays could serve as noninvasive surrogates for TME assessment (22, 23).
In this study, we aimed to assess the expression patterns of RPA1, RFC1, and POLE in ccRCC and adjacent normal renal tissues using immunohistochemical analysis and explore their potential prognostic value. We evaluated correlations between protein immunoexpression, clinicopathological features, and preoperative blood parameters in patients with ccRCC. Additionally, we investigated the association of mRNA expression profiles of POLE, RPA1, RPA2, RPA3, RFC1, RFC2, RFC3, RFC4, and RFC5 with survival outcomes in this cancer. Complementing these analyses, we examined the relationships between RPA1, RFC1, and POLE and immune checkpoint molecules and immune cell infiltration using TIMER2.0 and the ConsensusTME algorithm and further evaluated the prognostic impact of immune cell levels on ccRCC patient survival.
Materials and Methods
Patients and collection of tissue samples. The present study was performed with the approval of the Bioethical Commission of the University of Warmia and Mazury in Olsztyn, Poland (no. 4/2010 and 21/2017), and informed consent was obtained from all participants included in the study.
Tissue samples were collected from patients who underwent radical nephrectomy due to kidney tumor between 2010 and 2014. Paired tissue samples of ccRCC and macroscopically unchanged part of the kidney were obtained shortly after surgical resection. Specimens for routine histopathological evaluation and immunohistochemistry were placed in 4% buffered (pH 7.0-7.4) formalin and further processed into a paraffin block. Tumor stage and nuclear grade were classified according to the eighth edition of the TNM classification and Fuhrman nuclear grading, respectively (24, 25). Hematoxylin and eosin (H&E) staining performed on 4 μm sections of collected tumor and corresponding non-cancerous kidney tissues were evaluated by a pathologist to confirm the presence of ccRCC or a cancer-free phenotype, respectively.
A total of 52 patients with ccRCC were included in the study (29 men and 23 women, with a mean age of 62.5 years, range=27-83 years). None of the patients had a history of cancer other than ccRCC nor received preoperative therapy. The clinicopathological characteristics, routine preoperative blood parameters, and survival outcomes of all evaluated patients were collected during the study and are presented in Table I. The median follow-up was 38.7 months.
Baseline characteristics of patients diagnosed with clear cell renal cell carcinoma (n=52).
Immunohistochemistry, and evaluation of RPA1, RFC1, and POLE immunoreactivity. RPA1, RFC1, and POLE immunostaining was performed on formalin-fixed, paraffin embedded paired tumor and non-cancerous kidney tissue sections of patients with ccRCC following previously established protocols (17). Rabbit monoclonal anti-human antibodies directed against RPA1 [diluted 1:800 in phosphate-buffered saline (PBS), #ab79398, Abcam, Cambridge, UK], RFC1 (diluted 1:800 in PBS, #ab180613, Abcam), and POLE (diluted 1:400 in PBS, #ab110876, Abcam) were applied, while negative control slides were achieved by omitting the primary antibody. All immunostained sections of ccRCC and corresponding unchanged kidney were evaluated by two pathologists, independently (H.M. and A.S., coauthors of this study), blinded to clinical data. The immunoreactivity levels of RPA1, RFC1, and POLE were evaluated in tumor and unaltered renal tissue of patients with ccRCC according to the 12-point immunoreactive score system (IRS) of Remmele and Stegner (26). IRS is based on multiplication of the percentage of cells showing positive reaction (1-10%, 11-50%, 51-80%, >80% cells scored as 1, 2, 3, and 4 points, respectively) and reaction intensity (low, moderate, and intense reaction scored as 1, 2, and 3 points, correspondingly).
Bioinformatic analysis of RNA-Seq data. RNA-Seq gene expression data was obtained from the TCGA-KIRC cohort. Samples with “Primary tumor” status and a primary diagnosis of “Clear cell adenocarcinoma” were included, yielding 516 samples. Gene expression data for nine specified genes encoding POLE, subunits of RFC (RFC1, RFC2, RFC3, RFC4, RFC5), and RPA (RPA1, RPA2, RPA3) proteins were selected for further analysis. Raw read counts were transformed using variance stabilizing transformation to minimize heteroscedasticity, and the values were standardized to z-scores (27). Based on the normalized gene expression values of each gene, the dataset was divided into two groups: ‘Low’ and ‘High’ gene expression. These ‘Low’ and ‘High’ expression groups, along with their clinical information, were subjected to Kaplan-Meier survival analysis (28).
TIMER2.0 analysis. To evaluate the association between gene expression (RPA1, RFC1, and POLE) and immune cell infiltration levels in ccRCC, we used the TIMER2.0 tool (29). TIMER2.0 estimates the abundance of six immune cell types – B cells, CD4+ T cells, CD8+ T cells, macrophages, neutrophils, and dendritic cells – within the TME using bulk RNA sequencing data. The tool applies a deconvolution algorithm to RNA-Seq data to calculate infiltration levels and correlates these with gene expression. Moreover, we examined the association between RPA1, RFC1, and POLE and the most important immune-checkpoint molecules using this tool.
ConsensusTME analysis. To complement the TIMER2.0 analysis, we estimated the relative abundance of 18 cell types [including 16 immune cell subsets, endothelial cells, and cancer-associated fibroblasts (CAFs)] within the TME using the ConsensusTME algorithm (30). Bulk RNA-Seq data from ccRCC samples were processed, and cell type abundances were stratified according to RPA1, RFC1, and POLE expression (dividing samples into high and low groups using the median). Furthermore, to assess the prognostic impact of immune cell infiltration, patients with ccRCC with available immune subpopulation data (n=525) were stratified into subgroups based on the relative abundance (dichotomized as low versus high, using the median as a cutoff) of various immune and stromal cell types.
Statistical analysis. Statistical analyses were carried out using Statistica 13.1 (Statsoft, Tulsa, OK, USA) and R 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria) with the stats package (31), unless otherwise noted. To assess the lack of incidental significant association between study groups in terms of age and sex, Fisher’s exact test was conducted. The Wilcoxon test was used for comparing protein expression between specified categories. Cox regression analysis and Kaplan-Meier survival plots were performed using the R survival package (30) to assess the impact of gene expression, clinicopathological features, and preoperative blood parameters on OS. Two grouping strategies were employed for protein expression: 1) grouping based on the IRS scale (1-6 and 7-12 groups) and 2) a dichotomous division comparing IRS levels in tumor tissue to those in adjacent non-cancerous tissue for each patient (for instance, classifying them as having down-regulated versus unchanged/up-regulated RFC1 expression). Spearman’s rank correlation was utilized to examine the relationships between RPA1, RFC1, and POLE expression and clinicopathological features of patients with ccRCC.
For the TCGA cohort, normalized RNA-Seq gene expression data were dichotomized into “Low” and “High” groups based on z-scores, and survival outcomes were analyzed using Kaplan–Meier curves and Cox regression tests. For TIMER2.0 analysis, Spearman’s rank correlations (adjusted for tumor purity) were used to assess the relationships between RPA1, RFC1, and POLE expression and immune cell infiltration in ccRCC. In the ConsensusTME analysis, comparisons of immune cell abundances between high- and low-expression groups were carried out using t-tests. Cox regression and Kaplan–Meier survival analyses were then performed on the TCGA ccRCC cohort after stratifying patients into subgroups based on the relative abundance (low versus high) of specific immune and stromal cell types determined according to RPA1, RFC1, and POLE expression.
Results
RPA1 immunoexpression is elevated while RFC1 immunoreactivity is decreased in ccRCC cells. ccRCC cells exhibited mainly nuclear RPA1 protein expression (Figure 1A). In our cohort of patients with ccRCC, the median RPA1 immunoreactivity was significantly increased in cancer cells as compared to the corresponding normal renal tissues [median (IQR) IRS 2 (1-4) and 1 (1-2), respectively; p<0.001] (Figure 1B). In detail, RPA1 protein expression level was up-regulated in 29/52 (56%), remained unchanged in 20/52 (38%), and was down-regulated in 3/52 (6%) renal tumor sections compared with the matched unchanged kidney tissue.
Immunohistochemical expression and immunoreactive scores of RPA1, RFC1, and POLE in clear cell renal cell carcinoma (ccRCC) and adjacent normal tissues (n=52). (A) Representative immunohistochemical staining images of RPA1, RFC1, and POLE in ccRCC and matched adjacent normal kidney tissues. Scale bar: 40 μm. (B) Comparison of immunoreactive score (IRS) for POLE, RFC1 and RPA1 in healthy kidney tissue and clear cell renal cell carcinoma groups. The horizontal line inside the boxplots shows the median and the whiskers spread +/− 1.5 of interquartile range. Dots represent single observations.
RFC1 immunoreactivity was observed in both cytoplasm and nucleus of ccRCC and matched healthy tissue of kidney (Figure 1A). The median immunoexpression of RFC1 was significantly decreased in ccRCC compared with the matched unchanged renal tissue [median (IQR) IRS 2 (1-4) and 4 (2.5-8), respectively; p<0.001] (Figure 1B). More than half of patients with ccRCC had lower RFC1 protein levels in cancer cells compared with the corresponding non-cancerous kidney tissue (28/49, 57.1%), whereas RFC1 immunoexpression remained unchanged or elevated in 10/49 (20.4%) and 11/49 (22.4%), respectively.
POLE immunoreactivity was similar in ccRCC and non-cancerous renal tissue [median (IQR) IRS 2 (1-4) and 2 (1-4), respectively; p=0.239] (Figure 1B). POLE immunoexpression, localized predominantly in the cytoplasm (Figure 1A), was higher in 23/52 (44.2%), remained unchanged in 9/52 (17.3%), and was lower in 20/52 (38.5%) cancerous cells compared with the matched healthy kidney tissue.
RPA1, RFC1, and POLE expression in cancerous and tumor-adjacent tissues of ccRCC and correlation with clinicopathological factors. RPA1, RFC1, and POLE immunoexpression levels showed modest correlation with most demographic and clinicopathological factors in patients with ccRCC. However, several significant associations with preoperative blood parameters and tumor characteristics were observed (Table II, Figure 2A and B).
Correlations between protein immunoexpression levels (RPA1, RFC1, POLE) and clinicopathological factors as well as preoperative blood parameters in clear cell renal cell carcinoma (n=52).
Correlations between protein immunoexpression levels (RPA1, RFC1, POLE), clinicopathological factors, and preoperative blood parameters in our cohort of patients with clear cell renal cell carcinoma (n=52). Figure 2A presents correlations among protein immunoexpression levels (RPA1 IRS, RFC1 IRS, POLE IRS). Figure 2B shows correlations between protein immunoexpression levels and clinicopathological factors as well as preoperative blood parameters. The Spearman correlation coefficient (r) and corresponding p-values are reported. CREA: Creatinine; G group: grade group; HGB: hemoglobin; IRS: immunoreactivity score; LMR: lymphocyte-to-monocyte ratio; LYMPH: lymphocytes; MCV: mean corpuscular volume; MONO: monocytes; NEU: neutrophils; NLR: neutrophil-to-lymphocyte ratio; PLR: platelet-to-lymphocyte ratio; PLT: platelets; RBC: red blood cells; SII: systemic immune-inflammation index; T stage: tumor stage; WBC: white blood cells.
Increased RPA1 immunoexpression in cancerous cells of patients with ccRCC correlated with lower preoperative white blood count (WBC) (R=−0.333, p=0.019) and neutrophil levels (R=−0.366, p=0.010) (Table II). Additionally, both the neutrophil-to-lymphocyte ratio (NLR) (R=−0.448, p=0.001) and the systemic immune-inflammation index (SII) (R=−0.360, p=0.011) inversely correlated with RPA1 levels in tumor cells, hinting at a link between immune response and RPA1 expression (Table II). A positive correlation between RPA1 and RFC1 expression levels within cancerous cells was also observed (R=0.311, p=0.031) (Figure 2A).
As we observed that RFC1 protein expression in tumor cells inversely correlated with tumor size in our cohort (R=−0.303, p=0.032) (Table II), we further investigated its association with tumor dimensions. Patients with down-regulated RFC1 immunoexpression in cancer cells compared to adjacent normal tissue had significantly larger tumors than those with unchanged or up-regulated RFC1 levels [median (IQR): 8.0 cm (5.7-9.8) vs. 4.5 cm (3.5-6.3); p=0.00021]. Similarly, more patients with decreased RFC1 in cancer cells compared to adjacent non-cancerous tissue had tumors >7 cm (54% vs. 20%). Additionally, the negative associations between RFC1 levels and WBC (R=−0.268, p=0.063) as well as neutrophil counts (R=−0.276, p=0.055) were close to being significant (Table II).
In our cohort, POLE immunoreactivity inversely correlated with preoperative the platelet-to-lymphocyte ratio (PLR) (R=−0.280, p=0.049) (Table II). Notably, patients with platelet (PLT) values below the median (238 000/mm3) showed higher POLE expression compared to those with PLT values at or above the median [median (IQR)=4 (2-6) vs. 2 (1-3); p=0.033]. Furthermore, while not statistically significant, POLE level showed a trend towards an inverse correlation with tumor T stage (R=−0.244, p=0.084) (Table II), potentially linking POLE with early-stage disease.
In patients with larger tumor sizes, positive correlation was noted with PLT (R=0.276, p=0.050) (Figure 2B). Advanced T stage (T2/T3) positively correlated with PLR (R=0.387, p=0.005), PLT (R=0.456, p<0.001), and SII counts (R=0.296, p=0.035), while inversely correlating with hemoglobin levels (R=−0.418, p=0.002) (Figure 2B), further underscoring the role of systemic inflammation and anemia in tumor progression.
Several significant correlations were noted between grade and hematological parameters. Higher grades (G3/G4) were associated with decreased lymphocyte counts (R=−0.293, p=0.037) and LMR (R=−0.284, p=0.044), as well as elevated PLT (R=0.322, p=0.021) and platelet-to-lymphocyte ratio (PLR) (R=0.343, p=0.014) (Figure 2B). These findings emphasize the role of systemic inflammation and immune suppression in ccRCC patients with higher-grade tumors.
In ccRCC patients with metastases, elevated PLT and SII levels were observed (R=0.295, p=0.036 and R=0.289, p=0.039, respectively) (Figure 2B).
Survival analysis of POLE, RFC and RPA in ccRCC. Survival rates in ccRCC were significantly correlated with tumor stage (p=0.004), histologic grade (p=0.008), tumor size >7 cm (p=0.032), presence of distant metastasis (p=0.004), neutrophils (NEU)(p=0.022), red blood cells (RBC) (p=0.046), and WBC (p=0.004) counts, hemoglobin levels (p=0.008), as well as SII (p=0.005) in the univariate analysis (Figure 3A). However, in the multivariate analysis, only the presence of metastases (p=0.041) significantly affected survival in our cohort (Figure 3B). Sex and age did not correlate with survival rates of our cohort. These findings indicate that the study cohort is representative, as the identified prognostic markers are well-documented indicators of poor prognosis in ccRCC. The expression levels of POLE, RFC1, and RPA1 did not significantly correlate with survival in either analysis. Notably, relative RFC1 expression (cancer vs. normal adjacent renal tissue) was closest to statistical significance [HR=2.41 (0.87-6.70), p=0.091].
Univariate (A) and multivariate (B) analysis of patients with ccRCC (n=52). For preoperative blood variables, patients were divided into two groups based on the median values. Bold values indicate p<0.05. HGB: Hemoglobin; HR: hazard ratio; NEU: neutrophils; OS: overall survival; RBC: red blood cells; SII: systemic immune-inflammation index; WBC: white blood cells.
Some of investigated proteins consist of several subunits coded by specific genes: the POLE protein is encoded by the POLE gene, functioning as part of a four-subunit complex, whereas RFC protein is composed of five subunits encoded by RFC1, RFC2, RFC3, RFC4, and RFC5 genes. Also, RPA protein consists of three subunits encoded by RPA1, RPA2, and RPA3 genes. We investigated whether the expression level of these genes could predict the clinical outcome using the TCGA cohort of patients with ccRCC. Survival analysis was performed for each gene on 516 subjects (Stage I - 256, stage II - 55, stage III - 122, and stage IV - 83). Only RPA1, RFC1, RFC4, and POLE expression patterns correlated with survival of the ccRCC cohort. In detail, low expression of RPA1 and RFC1 correlated with worse prognosis. In contrast, ccRCC patients with low expression of RFC4 and POLE had better prognosis (Figure 4).
Kaplan-Meier survival curves showing the prognostic impact of POLE, RFC1-5, and RPA1-3 mRNA expression in patients with ccRCC from the TCGA cohort of 516 patients with ccRCC stratified by low and high mRNA expression levels. Significant p-values (<0.05) from corresponding Cox regression tests are indicated in bold.
Relationship between RPA1, RFC1, and POLE expression and immune cell infiltration in ccRCC. Given the critical role of various tumor-infiltrating immune cells in tumor progression and prognosis, we explored the association of RPA1, RFC1, and POLE expression and immune cell infiltration in ccRCC using the TIMER2.0 tool (Figure 5A). The results showed that both RPA1 and RFC1 expression were positively correlated with the infiltration levels of CD4+ T cells (R=0.202 and R=0.284, respectively, p<0.001), macrophages (R=0.367 and R=0.376, respectively, p<0.001), myeloid dendritic cells (R=0.246, p<0.001 and R=0.139, p=0.003, respectively), and neutrophils (R=0.454 and R=0.369, respectively, p<0.001). In addition, POLE expression was associated with CD4+ T cells and neutrophils (R=0.391 and R=0.206, respectively, p<0.001).
Immune infiltration in clear cell renal cell carcinoma based on RPA1, RFC1, and POLE expression. (A) TIMER2.0 analysis, after purity adjustment, reveals the association between RPA1, RFC1, and POLE expression and the infiltration levels of key immune cell populations in ccRCC. (B) Violin plots, visualized using the SRplot online tool, display the differential relative abundance of 18 tumor microenvironment cell types – including 16 immune-related cells, endothelial cells, and cancer-associated fibroblasts – in ccRCC, stratified by high (orange) and low (blue) expression of RPA1, RFC1, and POLE. Stars above the bars indicate a relatively higher abundance in the high expression group, whereas stars below indicate a higher abundance in the low expression group. Statistical significance is denoted as follows: p<0.05 (*), p<0.01 (**), p<0.001 (***), and p<0.0001 (****).
We also examined the correlation between the most significant immune-checkpoint molecules and RPA1, RFC1, and POLE (Figure 6). RPA1 expression was most prominently associated with PD-L1 (encoded by the CD274 gene) and VISTA (encoded by the VSIR gene), whereas RFC1 showed positive associations with PD-L1, VISTA, as well as a negative correlation with GITR (encoded by the TNFRSF18 gene). Moreover, POLE expression was positively linked to CTLA4 and GITR.
Correlations between RPA1, RFC1, POLE levels and immune checkpoint molecules – purity-adjusted TIMER2.0 analysis.
To better understand the impact of RPA1, RFC1, and POLE on immune cell infiltration, we analyzed bulk RNA-Seq data using the ConsensusTME algorithm (30) to estimate the relative abundance of 18 different cell types, including endothelial cells and cancer-associated fibroblasts (CAFs), as well as 16 immune-related cells in the tumor immune microenvironment of ccRCC (Figure 5B).
All three genes - RPA1, RFC1, and POLE - exhibited common patterns with key immune cells, particularly B cells, mast cells, M1 macrophages, and monocytes. Tumors with elevated expression of these genes had significantly higher relative levels of M1 macrophages and monocytes, as well as lower levels of B cells (p<0.05 for each). Notably, mast cell abundance differed: while high RPA1 and RFC1 expression correlated with increased mast cells, POLElow tumors had higher mast cell levels.
Beyond these shared patterns, gene-specific differences were evident. RPA1high tumors were marked by increased M1 macrophages (p<0.0001), neutrophils (p<0.0001), dendritic cells (p<0.01), mast cells (p<0.01), and monocytes (p<0.05) compared to RPA1low tumors. In contrast, B cell concentration was higher in the RPA1low cancers (p<0.05).
RFC1high tumors displayed considerably higher levels of dendritic cells, M1 macrophages, naïve macrophages (all three p<0.0001), neutrophils, monocytes, mast cells (all three p<0.01), endothelial cells, and M2 macrophages (both p<0.05). Contrarily, RFC1low tumors demonstrated greater prevalence of cytotoxic cells (p<0.0001), plasma cells, B cells (both p<0.001), natural killer (NK) cells, regulatory T cells, CD8+ T cells (all three p<0.01), CD4+ T cells, and gamma-delta T cells (both p<0.05).
ccRCC patients with increased POLE expression had higher relative concentrations of CD8+ T cells (p<0.001), gamma-delta T cells, regulatory T cells, monocytes, NK cells, CD4+ T cells (all five p<0.01), cytotoxic cells, and M1 macrophages (both p<0.05). Whereas, POLElow patients had increased levels of B cells (p<0.0001), CAFs (p<0.001), and mast cells (p<0.01).
Prognostic significance of immune cell infiltration in ccRCC according to RPA1, RFC1, and POLE expression. To assess the prognostic impact of immune cell infiltration in relation to RPA1, RFC1, and POLE expression, patients with ccRCC from the TCGA database (with available immune subpopulation data) were stratified into subgroups based on the relative abundance (dichotomized as low versus high using the median as a cutoff) of various immune and stromal cell types (Figure 7).
Forest plots showing the prognostic value of POLE, RPA1, and RFC1 mRNA expression in patients with clear cell renal cell carcinoma from TCGA database (n=525) stratified by abundance (HIGH or LOW) of immune cell types in the tumor microenvironment. Forest plots illustrate univariate Cox regression results assessing the prognostic value of the relative abundance of 18 tumor microenvironment cell types in ccRCC, stratified by RPA1, RFC1, and POLE mRNA expression. Patients were grouped into subgroups based on the relative abundance (dichotomized as low versus high using the median as a cutoff) of various immune and stromal cell types. The plots depict hazard ratios as squares, with horizontal lines (red for low and grey for high expression) representing 95% confidence intervals.
For RPA1, higher expression was associated with improved OS (HR=0.83, 0.72-0.96, p<0.01). In subgroup analyses, RPA1high patients with low abundance of B cells, dendritic cells, eosinophils, mast cells, neutrophils, and M1 macrophages exhibited significantly lower HRs (ranging from 0.73 to 0.78, all p<0.01), indicating better survival outcomes. Conversely, a high abundance of NK cells and fibroblasts (HR=0.74 and 0.75, respectively, both p<0.01) was also associated with improved OS.
For RFC1, higher expression significantly correlated with better OS (HR=0.80, 0.70-0.92, p=0.00163). In subgroup analyses, RFC1high patients with low abundance of B cells, eosinophils, neutrophils, and M1 macrophages (HRs ranging from 0.66 to 0.72, all p<0.001), along with high levels of cytotoxic cells, NK cells, endothelial cells, and fibroblasts (HR=0.69-0.73, all p<0.001) were associated with improved outcomes.
In contrast, higher POLE expression was significantly associated with poorer OS (HR=1.68, 1.43-1.98, p<0.000001). In POLEhigh patients, low abundance of B cells, eosinophils, M1 macrophages, and plasma cells (HR=2.11, 2.08, 2.02, and 2.02, respectively; all p<0.000001) as well as high abundance of cytotoxic cells and endothelial cells (HR=1.82 and 2.03, both p<0.000001) were associated with markedly poorer survival.
Notably, across all three genes, the abundance of B cells, eosinophils, and M1 macrophages consistently impacted survival outcomes, underscoring their potential as key modulators of the tumor immune microenvironment in ccRCC.
Discussion
In the present study, we compared ccRCC tissues with adjacent normal renal tissues using IHC and observed distinct expression patterns for RPA1, RFC1, and POLE. Specifically, RPA1 was significantly up-regulated in tumor cells, RFC1 was down-regulated, and POLE immunoexpression remained largely unchanged.
The observed overexpression of RPA1 aligns with its role in maintaining genome integrity during replication stress (32, 33). In several malignancies – including gastrointestinal, bladder, and brain cancers (34-39) – elevated RPA1 suggests a compensatory response to increased replication stress, although, its down-regulation in early breast tumorigenesis highlight its complex, context-dependent function (40). In our cohort, higher RPA1 levels were linked to lower preoperative WBC and neutrophil counts, as well as reduced NLR and SII, suggesting a less activated systemic inflammatory profile. Furthermore, we observed the positive correlation between RPA1 and RFC1 expression, hinting at a coordinated regulation of these replication factors in ccRCC. Sejima et al. linked low NLR with increased Fas ligand expression and improved survival in RCC (41), potentially via cytotoxic T cell apoptosis (42). Similarly, our finding that higher RPA1 expression aligns with lower NLR may reflect an immune milieu favoring anti-tumor responses. Furthermore, the TIMER2.0 analysis revealed that tumors with high RPA1 expression had increased local infiltration of CD4+ T cells, macrophages, myeloid dendritic cells, and neutrophils. Given that neutrophils can polarize into either pro-tumor (N2) or anti-tumor (N1) phenotypes (43) – with N1 neutrophils capable of mediating Fas ligand-dependent apoptosis (44) – the association of high RPA1 with improved prognosis in TCGA ccRCC data may indicate a TME enriched in anti-tumor N1 neutrophils.
RFC1 was significantly down-regulated in our ccRCC samples. As a key subunit of the pentameric RFC complex responsible for PCNA loading during DNA replication and repair (45), its reduced expression may reflect impaired replication and repair processes, contributing to genomic instability and tumor progression.
Notably, lower RFC1 levels were associated with larger tumor sizes (median 8.0 cm versus 4.5 cm), and, consistent with RPA1 findings, higher RFC1 immunoexpression trended with reduced preoperative WBC and neutrophil counts. TIMER2.0 analysis mirrored the immune associations seen with RPA1, as higher RFC1 expression correlated with increased infiltration of CD4+ T cells, macrophages, myeloid dendritic cells, and neutrophils. Moreover, TCGA data identified low RFC1 as a negative prognostic marker in ccRCC. Further investigation is warranted to elucidate the mechanisms driving RFC1 suppression and its broader implications for ccRCC biology.
Although POLE immunoreactivity did not differ significantly between tumor and normal tissues in our study, further analysis uncovered clinically meaningful correlations. Elevated POLE levels were associated with lower preoperative PLR and an inverse correlation with T stage, implying that robust POLE expression may characterize early-stage ccRCC. In fact, patients with platelet counts below the median (238,000/mm3) exhibited significantly higher POLE scores. TIMER2.0 data also linked POLE expression with increased CD4+ T cell and neutrophil infiltration, suggesting that POLE might modulate the TME in a stage-dependent manner. Wu et al. and our TCGA analysis revealed that elevated POLE is associated with poorer overall survival in ccRCC (15). However, it is possible that in early tumorigenesis, high POLE expression supports active replication and repair mechanisms, while its reduction in advanced stages may reflect a less proliferative state or improved genomic stability, ultimately correlating with better outcomes.
Our results highlight complex interactions between replication protein expression and systemic inflammatory markers in ccRCC. In our cohort, elevated RPA1 correlated with lower preoperative WBC, neutrophils, NLR, and SII, with RFC1 showing similar near-significant trends. POLE expression inversely correlated with PLR, and patients with lower platelet counts exhibited higher POLE immunoreactivity. Moreover, advanced tumor features – including larger size, higher T stage, grade, and the presence of metastasis – were accompanied by elevated PLR, NLR, and SII, and decreased LMR. These patterns suggest that immune responses interact with DNA replication and repair pathways to shape tumor progression and potentially affect treatment response in ccRCC.
In addition, analysis of TCGA ccRCC cohort, which included POLE and all subunits of the RFC and RPA at the mRNA level, provided further prognostic insights. Specifically, low expression of RPA1 and RFC1 was linked to worse prognosis, while lower expression of RFC4 and POLE was associated with improved survival. The contrasting prognostic roles within the RFC complex suggest that individual subunits may have distinct functions beyond DNA replication (46-49). Immunological assessments using TIMER2.0 indicated that while RPA1 and RFC1 expression positively correlated with macrophage and neutrophil infiltration, POLE was predominantly associated with CD4+ T cell presence. ConsensusTME analysis further demonstrated that tumors with high expression of these genes generally had more M1 macrophages and monocytes but fewer B cells. Subgroup survival analyses revealed that, in the context of high RPA1 or RFC1 expression, lower abundances of B cells, eosinophils, neutrophils, and M1 macrophages - as well as higher levels of NK cells and CAFs - predicted improved OS. In contrast, high POLE expression was associated with markedly poorer outcomes, particularly when coupled with low levels of B cells, plasma cells, eosinophils, and M1 macrophages, and high endothelial cell infiltration. These data highlight the functional heterogeneity of immune cell subsets within the TME and their variable prognostic impacts. For instance, in high RPA1 or RFC1 tumors, exclusion of B cells – normally immunosuppressive (50, 51) – was linked to better survival, whereas low B cell abundance in high POLE tumors predicted worse outcomes, suggesting a potentially anti-tumor role for B cells. Similarly, lower eosinophil levels in high RPA1 or RFC1 tumors associated with better survival, which could be explained by a reduction in subsets of tumor-promoting eosinophils, while in high POLE tumors, low eosinophil levels correlated with poorer outcomes, indicating eosinophils may support anti-tumor immunity (52). These findings emphasize that the prognostic impact of immune cell infiltration is context-dependent, influenced by the tumor’s molecular profile and its immune environment, which could affect both progression and treatment response.
Our study also revealed possible interactions between replication proteins and immune checkpoint molecules. POLE expression was notably linked to CTLA-4 and GITR, whereas RPA1 and RFC1 were associated with PD-L1 and VISTA, and only RFC1 negatively correlated with GITR. Although in previous studies high POLE expression has been connected to an immunosuppressive TME and elevated immune-checkpoint molecules (PD-1, CTLA-4, CD86), suggesting that POLE may promote immune evasion and resistance to ICIs (15), increased checkpoint molecule expression can paradoxically enhance ICI responsiveness (53). As POLE mutations are rare in ccRCC – limiting their utility as predictive biomarkers – the assessment of POLE expression may offer more practical insights into tumor immune evasion and response to immunotherapy (13, 15). Given the challenges of resistance to first-generation ICIs (targeting CTLA-4, PD-1, and PD-L1 inhibitory molecules) and the emergence of second-generation modulators (such as those directed against LAG-3 molecule - relatlimab - and agents targeting GITR or VISTA) (54, 55), our findings underscore the potential of targeting DNA replication/repair pathways in combination with immunotherapy.
While our study benefits from an integrated approach combining tissue-based IHC with TCGA transcriptomics - it is not without limitations. The relatively small sample size of our cohort, focus solely on mRNA and protein levels (without accompanying mutation or functional analyses), and the retrospective single-center design may limit the generalizability of our results.
Conclusion
In summary, our study reveals distinct expression profiles of RPA1, RFC1, and POLE in ccRCC and highlights their interplay with systemic inflammation and the tumor immune microenvironment. The integration of IHC data from our patient cohort with transcriptomics and immune profiling from TCGA underscores the clinical and immunological relevance of replication proteins in disease progression. These findings suggest that targeting DNA replication and repair pathways may enhance therapeutic strategies, especially in combination with immune checkpoint inhibitors. Further research is warranted to validate these associations in larger, prospective cohorts and to explore their mechanistic underpinnings.
Acknowledgements
The Authors would like to thank Anna Koprowicz-Wielguszewska, M.Sc. for technical help and Kamil Myszczyński, PhD for the TCGA datasets’ extraction.
Footnotes
Authors’ Contributions
MG, JG, BEK and JK conceived the study and reviewed the manuscript. MG, JK, HM, and AS performed experiments, collected and analyzed data. MG wrote the manuscript. JK, BEK, HM, and AS confirm the authenticity of all the raw data. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no competing interests in relation to this study.
Funding
This study was supported by a statutory grant of the School of Medicine, Collegium Medicum, The University of Warmia and Mazury in Olsztyn, Poland.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received September 19, 2025.
- Revision received October 2, 2025.
- Accepted October 3, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.