In this study, differential analysis was conducted utilizing the mRNA sequencing data from the TCGA-KIRC dataset, which includes 72 normal samples and 532 KIRC samples ( Supplementary Table S1 ), and 32 KIRC and 23 normal samples from the GSE15641 dataset ( Supplementary Table S2 ) in the Gene Expression Omnibus (GEO) database. The clinical information of 39 KIRC samples in the GSE29609 ( Supplementary Table S3 ) dataset was used for WGCNA analysis. From January 2024 to April 2024, 20 pairs of cancerous and nearby normal tissues from KIRC patients were surgically removed at Soochow University’s First Affiliated Hospital. Postoperative pathology verified the specimens. Prior to the procedure, anti-tumor treatment was not administered to any of the patients. The Ethics Committee of Soochow University’s First Affiliated Hospital granted approval for this study under the number 2024-395. Written informed consent forms were signed by each patient.

The investigation of mRNA differential expression was conducted using the R software’s Limma package (version 3.40.2). Differential analysis was carried out using the GSE15641 and TCGA-KIRC datasets, and the screening criteria were Log2 (Fold Change) > 2 or Log2 (Fold Change) < -2 and P < 0.05 ( Supplementary Figures S1A, B ). In this study, all selected genes were upregulated. The GSE29609 dataset, which includes prognostic information for 39 cases of KIRC, was analyzed to identify the module with the highest correlation using WGCNA. To meet the assumption of a scale-free network as closely as possible, it was necessary to determine an appropriate value for the adjacency matrix weight parameter, power. The power value was set to range from 1 to 30, and the corresponding network correlation coefficients and mean connectivity were calculated. A higher correlation coefficient (with a maximum value of 1) indicates a closer fit to a scale-free network distribution. However, to ensure the robustness of the network, the power value was selected to balance a sufficiently high correlation coefficient with adequate gene connectivity. In this analysis, the power value was set to 7, as shown in Supplementary Figures S1C, D . Based on the selected power value, a weighted gene co-expression network model was constructed, resulting in the division of genes into six modules. The gray module represents a collection of genes that could not be assigned to any specific module ( Supplementary Figure S1E ). Among the modules, the brown module exhibited the highest correlation, with a correlation coefficient of 0.34 ( Supplementary Figure S1F ).

The data were examined using functional enrichment to further validate the possible roles of the possible targets. A popular technique for annotating genes with functions is Gene Ontology (GO), particularly for molecular function (MF), biological process (BP), and cellular component (CC). Gene functions and related advanced genomic functional information can be analyzed with the use of the useful Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis. The ClusterProfiler package in R was used for GO and KEGG enrichment analysis, as well as to investigate possible aspects of Gene Set Enrichment Analysis (GSEA), in order to better understand the target gene’s carcinogenic role (24). To control for false positives resulting from multiple testing, we applied the Benjamini-Hochberg method to adjust the p-values, calculating the false discovery rate (FDR). Enriched terms with an adjusted p-value (FDR) of less than 0.05 were considered statistically significant.

In order to guarantee a trustworthy assessment of the immune score outcomes, we utilized the R software package immunedeconv. Every algorithm had distinct benefits and had undergone extensive testing. The XCELL approach was used for this investigation since it evaluates a greater variety of immune cells. Additionally, we identified immune cells with prognostic value using the LASSO algorithm. R Foundation for Statistical Computing version 4.0.3 was used to implement all of the previously discussed analytic techniques and R packages.

The ENCORI and TarBase v.8 databases were used to analyze PAQR6-related miRNA. The ENCORI database was used to analyze circRNA related to miRNA (25, 26).

769P cell was cultured in RPMI 1640 medium (cytiva, UK) containing 10% fetal bovine serum (Gibco, USA) and the cell was cultured in a 37°C cell incubator containing 5% CO2. PAQR6 knockdown lentivirus were purchased from Shanghai Genechem (China). Viral infection was performed according to the instructions, and then screening was performed with puromycin.

RIPA lysate (Beyotime, China) combined with a proteinase inhibitor was used to extract the proteins. After loading proteins onto an SDS-PAGE gel, they were moved onto a PVDF membrane. After blocking the membrane with 5% skim milk for two hours at room temperature, the matching primary antibody (abs143408, absin) was incubated at 4°C for the entire night. Lastly, an enhanced chemiluminescence (ECL) kit (Beyotime) was used to expose the membrane after it had been treated with secondary antibodies coupled with horseradish peroxidase.

The supplier of the EdU assay kit was Beyotime Company (C0078S, China). Transfected cells were exposed to EdU reagent for two hours in accordance with the manufacturer’s instructions. The cells were then incubated with 0.3% Triton X-100 for 15 minutes at room temperature after being fixed with 4% paraformaldehyde for 15 minutes. Lastly, cells were stained using Hoechst and fluorescent dye. The pictures were taken with a Nikon TI2-D-PD inverted microscope (Japan).

Following the creation of stably transfected cell lines, the cell plate was scratched with 200 μl of Eppen-dorf Tip, and the cells were then rinsed two or three times with PBS. 1% fetal bovine serum is still used to cultivate cells (Gibco, USA). Use an inverted microscope to see how the cells in each plate change at 0 and 12 hours.

A coating of Matrigel matrix glue (Corning, USA) (matrix glue: serum-free medium=1:6) was applied to the upper chamber. Cells are resuspended in serum-free media 36 hours after transfection. Next, 500 μl of complete media was added to the lower chamber, and 5×104 cells were transferred to the upper chamber. Fix the cells in the upper chamber with 4% paraformaldehyde for 30 minutes at room temperature after 24 hours, and then stain them for 20 minutes with crystal violet. The pictures were taken with a Nikon TI2-D-PD inverted microscope (Japan).

The expression of PAQR6 in KIRC and normal kidney tissues was detected by the Wilcoxon rank-sum test. The log-rank test was used for prognostic analysis. It was deemed statistically significant when the P value was less than 0.05.

Two different datasets (TCGA-KIRC and GSE15641) were used to group by cancer and adjacent cancer tissues and then screened differentially expressed genes through differential analysis ( Supplementary Figures S1A, B ). Subsequently, based on WGCNA, we further screened the key genes related to the prognosis of KIRC ( Supplementary Figures S1C–F ). A Venn diagram was used to intersect the results from these analyses, leading to the identification of PAQR6 as the key gene for subsequent investigations ( Figure 1A ). According to our findings, PAQR6 was consistently upregulated in KIRC samples as compared to normal renal tissues ( Figures 1B, C ).

To assess the prognostic implications of PAQR6, we analyzed data from the TCGA-KIRC dataset and observed that high PAQR6 expression was linked to lower rates of overall survival(OS) ( Figure 1D ). The finding was corroborated using the GSE29609 dataset ( Figure 1F ). To deepen our understanding, we performed disease-specific survival (DSS) and progression-free survival (PFS) analyses. DSS, which measures survival probability without death specifically attributable to KIRC, was analyzed using the TCGA-KIRC dataset ( Figure 1E ), providing insights into the relationship between PAQR6 expression and cancer-specific mortality. Similarly, PFS, assessed using the GSE29609 dataset ( Figure 1G ), evaluates the time to disease progression, including tumor recurrence or metastasis, offering critical insights into whether elevated PAQR6 expression is linked to more aggressive disease behavior or shorter recurrence-free intervals.

By combining DSS and PFS analyses, we provided a comprehensive evaluation of PAQR6’s prognostic significance, demonstrating its association with higher cancer-specific mortality and accelerated disease progression. In conclusion, our findings suggest that elevated PAQR6 expression in KIRC is strongly linked to poor patient prognosis, underscoring its potential as a prognostic biomarker.

In this study, we investigated the potential role of PAQR6 in KIRC. To identify the differentially expressed genes related to PAQR6, we applied screening criteria of P < 0.05, Log2 (Fold Change) > 2 or Log2 (Fold Change) < -2 ( Figure 2A ). The function of PAQR6 in KIRC was then further confirmed by functional enrichment analysis. KEGG pathway analysis revealed that upregulated genes were significantly enriched in pathways such as herpes simplex virus type 1 infection, Rap l signaling, phospholipase D signaling, and NOD-like receptor signaling ( Figure 2B ). These increased genes were implicated in RNA splicing, according to GO analysis, histone modification, covalent chromatin modification, and ciliary organization ( Figure 2C ). Conversely, KEGG analysis of downregulated genes identified their involvement in pathways like PI3K-Akt signaling, actin cytoskeleton regulation, cancer-related proteoglycans, and complement and coagulation cascades ( Figure 2D ). GO analysis revealed a connection between downregulated genes and extracellular matrix and structural organization, negative regulation of hydrolase activity, and glycosaminoglycan metabolic processes ( Figure 2E ).These findings indicate that PAQR6 plays a complex role in KIRC through multiple biological processes.

General differential analysis methods such as GO and KEGG may overlook genes with subtle expression differences that are still biologically significant. These methods do not fully account for important factors like gene regulator networks and the functional significance of gene interactions. To overcome these limitations, we performed a more thorough analysis of PAQR6 in KIRC using GSEA. Our findings revealed a significant correlation between PAQR6 and the immune microenvironment pathway of KIRC, including B cell receptor signaling, Toll-like receptor 1 and Toll-like receptor 2 cascades ( Figures 3A–C ). GSEA also confirmed a significant association between PAQR6 and pathways involved in angiogenesis and pluripotent stem cell differentiation ( Figures 3H, I ). Key target genes regulated by PAQR6, such as HIF1A, RAC1, EGFR, and IL1A, were identified ( Figures 3D–G ). Finally, the Gendoma database was used to identify the common interaction proteins between PAQR6 and these four target genes ( Figures 3J–M ). Protein interaction analysis further supported these associations, linking PAQR6 with angiogenesis and stem cell differentiation processes. In conclusion, the GSEA analysis provided detailed understanding of the multiple potential functions of PAQR6 in KIRC.

In this study, 38 immune cell types from the TCGA-KIRC dataset were evaluated using the XCELL method, identifying 12 that were significantly associated with KIRC prognosis. 5 key immune cell types were incorporated into a prognostic model: hematopoietic stem cells, monocytes, naive B cells, NK T cells, and CD4+Th1 T cells ( Supplementary Figures S2A–L , S3A, B ). A risk score formula based on the expression of these cells demonstrated a significantly poorer prognosis for high-risk patients, who also exhibited reduced responses to immune checkpoint inhibitors, as indicated by elevated TIDE scores ( Supplementary Figures S3C–E ). Hematopoietic stem cells, monocytes, and naive B cells were all validated by Cox regression analysis as possible prognostic indicators for KIRC ( Supplementary Figures S3F, G ). Further investigation showed that the expression of various immune regulatory genes varied significantly between the high-risk and low-risk groups, highlighting the distinct immunological profiles of these groups ( Supplementary Figures S4A–C ). The infiltration scores of 16 immune cell types showed notable differences between the groups with high and low PAQR6 expression, according to our analysis of the relationship between PAQR6 expression and immune cell infiltration in KIRC ( Figure 4A ). These results highlight PAQR6’s crucial function in the immune microenvironment of KIRC, with significant correlations to both immune stimulatory and suppressive factors, as well as immune cell infiltration.

Single-cell analysis was used to investigate the connection between PAQR6 expression and immune cell infiltration in KIRC. In the GSE111360 dataset ( Supplementary Table S4 ), PAQR6 expression was significantly correlated with various immune cell types, including CD4Tconv, Treg cells, Tprolif cells, CD8T cells, CD8Tex cells, NK cells, B cells, plasma cells, DC cells, monocytes/macrophages, mast cells, and fibroblasts ( Figures 4B, C ). A heatmap was generated to illustrate these correlations ( Figure 4D ). A correlation network graph was used to illustrate the substantial relationship between PAQR6 expression and immunological scores that was found through additional research utilizing the XCELL and TIP algorithms ( Figure 4E ).

We investigated the regulatory effects of PAQR6 on CD8+ T cells, which have been shown to have a prognostic significance. Our findings indicate that PAQR6 may affect KIRC prognosis via modifying CD8+ T cells. Additionally, we examined the relationship between PAQR6 expression and immune checkpoint genes, revealing notable distinctions between groups with high and low PAQR6 expression ( Figure 4F ). According to TIDE study, a poor prognosis after immune checkpoint inhibitor therapy is linked to increased PAQR6 expression ( Figure 4G ). These findings emphasize PAQR6’s role in shaping the immune microenvironment and its potential impact on immune-based therapies.

Using the ENCORI database, we identify 18 miRNAs with potential targeting relationship to PAQR6, while the miRWALK database identified 1,907 such miRNAs. From these datasets, we identified 14 miRNAs that were common to both ( Figure 5A ). Among these 14 miRNAs, 4 were found to have prognostic differences in KIRC ( Figure 5B ). We analyzed the correlation between the miRNAs and PAQR6 expression in KIRC, revealing a negative correlation between the miRNAs and their target gene. Specifically, hsa-miR-31-5p and hsa-miR-324-3p exhibited a strong negative correlation with PAQR6 in KIRC ( Figure 5C ). Normal kidney tissues had higher levels of hsa-miR-31-5p and hsa-miR-324-3p expression than KIRC specimens ( Figure 5D ).

After examining circRNAs that target these miRNAs in more detail, we discovered 40 circRNAs that were expressed differently in KIRC samples than in normal kidney samples ( Figure 5E ). Among these, 3 circRNAs were associated with KIRC prognosis. Sequence information for hsa-miR-31-5p, hsa-miR-324-3p, and the three prognostic circRNAs was provided ( Figures 5F, G ). The correlation analysis confirmed a negative relationship between these circRNAs and hsa-miR-31-5p/hsa-miR-324-3p in KIRC ( Figures 5H, I ). We investigated the cellular localization of PAQR6, hsa-miR-324-3p, hsa-miR-31-5p, and the three circRNAs, finding that PAQR6 is predominantly expressed on the plasma membrane ( Figures 5J, K ). Our analysis concluded that the prognosis of KIRC was connected with the ceRNA network including PAQR6-hsa-miR-31-5p/hsa-miR-324-3p-SH3BP2/GIGYF1/TRAK2/PRSS23. Finally, on the basis of the aforementioned research, we hypothesized the potential carcinogenic mechanism of PAQR6. The three circRNAs, SH3BP2, GIGYF1, and PRSS23, competitively bind to miR-31-5p and miR-324-3p, resulting in an elevation in the expression of PAQR6-related mRNA. Ultimately, this gives rise to the progression and invasion of KIRC ( Figure 5L ). Naturally, for a more precise carcinogenic mechanism, further investigations are requisite in the subsequent studies.

PAQR6 may be involved in the angiogenesis and pluripotent stem cell differentiation pathways in KIRC cells, according to gene enrichment analysis. Using the Genecards database, we identified 29 key prognostic genes related to these pathways ( Figure 6A ). Among these, the expression of PAQR6 and genes such as ACE, ANGPT2, BIRC5, CCND1, CDH5, CXCR4, EPO, FLT1, ICAM1, MCAM, PECAM1, SERPINE1, TGFB1, and VCAM1 showed no significant correlation between high- and low-PAQR6 expression groups. The strongest correlation was observed between PAQR6 and EZH2 ( Figures 6B–D ). Considering PAQR6’s potential oncogenic effects through interaction with EZH2, we tested the binding affinity of two EZH2 inhibitors, Tazemetostat and GSK2816126, which are currently in clinical trials. Strong binding to PAQR6 was demonstrated by both inhibitors ( Figure 6E ). EZH2, a critical transcription factor and well-known oncogene (27–29), plays a central role in regulating gene expression primarily through chromatin modifications. To further explore the transcriptional regulatory relationship between EZH2 and PAQR6, we performed ChIP-seq analysis. As shown in Figure 6F , significant EZH2 binding peaks were observed within the promoter and regulatory regions of PAQR6 under control (CTR) conditions. In contrast, treatment (None) significantly altered these binding patterns, indicating a dynamic interaction between EZH2 and PAQR6 under different conditions. These findings suggest that EZH2 may directly regulate PAQR6 expression through epigenetic mechanisms, potentially linking this interaction to angiogenesis- and stemness-related pathways, thereby contributing to tumor progression. To further evaluate PAQR6’s potential as a therapeutic target for KIRC, we tested four angiogenesis-related drugs (sunitinib, sorafenib, pazopanib, and axitinib). The robust binding of these drugs to PAQR6 ( Supplementary Figure S5 ) underscores its role in angiogenesis and its promise as a potential target for KIRC therapy.

Western blot was performed to evaluate the protein level of PAQR6 in normal renal cell (HK-2) and KIRC cell lines (786-0 and 769P) ( Figure 7A ). The expression level of PAQR6 was significantly increased in KIRC cell lines, particularly in 769P cell, compared with normal renal cell. Therefore,769P cell was selected to explore the role of PAQR6 in KIRC phenotypes. The knockdown treatment was performed using three different siRNAs. Compared to the control, the protein level of PAQR6 was significantly decreased in all three treatment groups ( Figure 7B ). The EdU assay was then performed to evaluate the impact of PAQR6 on tumor cell proliferation. In the knockdown group, the decreased PAQR6 significantly inhibited cell proliferation ( Figures 7C, D ). The scratch assay demonstrated that the knockdown of PAQR6 evidently inhibited the migration of 769P cell ( Figures 7E, F ). Moreover, the number of invading cells was significantly reduced compared to the control group, demonstrating that the knockdown of PAQR6 could inhibit cancer cell invasion ( Figures 7G, H ). These results indicated that the decreased PAQR6 could inhibit the proliferation, migration and invasion of KIRC cells. These findings supported the hypothesis that targeting PAQR6 could serve as a therapeutic strategy for inhibiting tumor growth and metastasis.