A Novel Cuproptosis-Related Prognostic Model and the Hub Gene FDX1 Predict the Prognosis and Correlate with Immune Infiltration in Clear Cell Renal Cell Carcinoma

Clear cell renal cell carcinoma (ccRCC) is a common malignancy of the urological system with poor prognosis. Cuproptosis is a recently discovered novel manner of cell death, and the hub gene FDX1 could promote cuproptosis. However, the potential roles of cuproptosis-related genes (CRGs) and FDX1 for predicting prognosis, the immune microenvironment, and therapeutic response have been poorly studied in ccRCC. In the present study, The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) data were downloaded. CRGs were subjected to prognosis analysis, and three of them were used to construct the prognostic model by least absolute shrinkage and selection operator (LASSO) regression. The CRGs prognostic model showed excellent performance. Moreover, based on the risk score of the model, the nomogram was developed to predict 1-year, 3-year, and 5-year survival. Furthermore, the hub gene of cuproptosis, FDX1, was an independent prognostic biomarker in multivariate Cox regression analysis. The pan-cancer analysis showed that FDX1 was significantly downregulated and closely related to prognosis in ccRCC among 33 cancer types. Lower FDX1 was also correlated with worse clinicopathologic features. The lower expression of FDX1 in ccRCC was verified in the external database and our own database, which may be caused by DNA methylation. We further demonstrated that the tumor mutational burden (TMB) and immune cell infiltration were related to the expression of FDX1. Immune response and drug sensitivity analysis revealed that immunotherapy or elesclomol may have a favorable treatment effect in the high FDX1 expression group and sunitinib or axitinib may work better in the low FDX1 expression group. In conclusion, we constructed a CRGs prognostic model and revealed that FDX1 could serve as a prognostic biomarker and predict therapeutic response in ccRCC. The study will provide a novel, precise, and individual treatment strategy for ccRCC patients.


Introduction
Renal cell carcinoma (RCC) ranks third among all genitourinary neoplasms, and most of them are clear cell renal cell carcinoma (ccRCC) [1,2]. Partial or radical nephrectomy were the primary treatment of localized ccRCC [3]. About 30% of ccRCC patients were found in advanced stages or presented with metastases at initial diagnosis [4,5]. In ccRCC, 70%∼90% of patients had VHL mutations, which immune inhibitors, challenges still remain for further improving prognosis because of drug resistance and adverse reactions [11,12]. Tere were some emerging therapeutic strategies, such as nanomedicine, that could overcome the heterogeneity of drug response and resistance of tumors [13]. Overall, the identifcation of new biomarkers is important not only for predicting prognosis but also for individual therapy.
Cuproptosis is a novel manner of cell death that was recently discovered and may provide a new target for cancer treatment [14]. As we all know, there were a great number of predetermined and precisely controlled programmed cell deaths throughout the development of multicellular organisms, such as apoptosis, necroptosis, pyroptosis, and ferroptosis [15][16][17]. In recent decades, we have also learned that various metals could cause cell death through other pathways than apoptosis. Zinc could trigger cell death by inhibiting adenosine triphosphate (ATP) synthesis [18]. Iron could trigger ferroptosis by catalyzing the formation of toxic membrane lipid peroxides [19]. Silver-based metal ions also had cytotoxic potential in various cancer cell lines through the induction of mitochondrial damage, oxidative stress, and autophagy [20]. As for copper, the study revealed that excess copper could perturb a set of lipoylated metabolic enzymes of the tricarboxylic acid (TCA) cycle and cause the loss of ironsulfur cluster proteins, leading to proteotoxic stress and ultimately cell death. Furthermore, the study showed a strong link between copper toxicity and mitochondrial activity. In ccRCC, HIF-1α could promote the shift of cellular metabolism away from the TCA cycle to glycolysis and downregulate mitochondrial respiration by regulating downstream genes such as pyruvate dehydrogenase and miR-210 [21,22]. Terefore, we speculated that there was a tight correlation between cuproptosis-related genes and ccRCC.
In the previous study, 10 genes were found to be closely related to cuproptosis. Among them, 7 genes (FDX1, LIAS, LIPT1, DLD, DLAT, PDHA1, and PDHB) could promote cuproptosis, and 3 genes (MTF1, GLS, and CDKN2A) could inhibit cuproptosis. FDX1, a reductase that reduces Cu 2+ to its more toxic form, Cu 1+ , was identifed as a hub gene that regulated protein lipoylation to promote cuproptosis. Previous studies demonstrated that the FDX1 protein was not detected in the lysates of most tissues, but it was highly expressed in the adrenal gland, kidney, and testes [23]. Tis result may indicate the unique role of FDX1 in kidney tissues and the development of ccRCC. Te deletion of FDX1 could cause resistance to cell death induced by one of the copper ionophores, elesclomol. In recent studies, elesclomol showed a hopeful result for the treatment of epithelial cancer, which indicated FDX1 may serve as a biomarker candidate [24][25][26]. However, its role has been rarely studied, especially in ccRCC.
In this study, we downloaded and analyzed the TCGA database to construct a cuproptosis-related prognostic model in ccRCC. Moreover, we identifed the hub gene FDX1 as being downregulated, especially in KIRC, and related to DNA methylation. We also detected the association between FDX1 and the immune microenvironment, and drug susceptibility to immunotherapy or target drugs. Tis study may provide an alternative model and reveal the important role of FDX1 in predicting prognosis and therapeutic efects in ccRCC.

Patients and Clinical Samples Collection.
A total of 38 ccRCC patients enrolled in this study signed informed consent, and this research was authorized by the Ethics Committee of Peking University First Hospital (Beijing, China). All of them underwent a partial or radical nephrectomy. Fresh tumor tissues and pair-matched adjacent normal tissues were obtained from those patients. All tissue samples were immediately stored in liquid nitrogen with RNAlater solution (Termo, AM7020, USA).

Diferentially Expressed Genes and Functional Enrichment
Analysis. Te diferentially expressed genes were analyzed using the "limma" R package. "Adjusted P < 0.05 and |Log 2 (fold change)| ≥ 1" were defned as the threshold for the diferential expression of mRNAs. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment were analyzed using the "ClusterProfler" R package (version: 3.18.0) [32].

Construction and Validation of the CRGs Prognostic
Model. Te patients in KIRC were randomly split into the training (n � 352) and validation (n � 178) cohorts at 2 : 1 ratio, and the clinical information is summarized in Table 1. Te 10 CRGs were subjected to univariate Cox regression analysis to fnd the genes, which could infuence overall survival (OS) and progression-free survival (PFS). Ten, there were three CRGs that were both signifcantly diferentially expressed and associated with prognosis. Tese genes were used to construct CRGs prognostic model using least absolute shrinkage and selector operation (LASSO) analysis in the training cohort. Te risk score was calculated as follows: Risk Score � β 1 x 1 + β 2 x 2 + β 3 x 3 +. . .+β n x n . Te patients were divided into two diferent groups according to the median risk score. Te prognostic model was assessed using Kaplan-Meier (KM) and time-dependent receiver operating characteristic (ROC) analysis by employing the R packages "survival" and "timeROC." A validation cohort was used to identify the performance of this model.

Establishment and Evaluation of the Nomograms.
We used univariate and multivariate Cox regression analyses to determine whether risk scores and other clinicopathological factors (age, gender, pathological T stage (pT), pathological M stage (pM), tumor stage, and histopathological grade) could be independent predictors of survival in ccRCC patients. Afterwards, we developed a nomogram that can assess the survival probability at 1, 3, and 5 years using all independent clinical prognostic factors. Calibration curves were drawn for 1, 3, and 5 years to compare the observed prediction probability with the actual OS probability.

Mutation Gene and Copy Number Variation (CNV)
Analysis. Te data of mutations were visualized using the "maftools" R package [36]. Using the mutations per million bases of each sample, we calculated the tumor mutational burden (TMB) value. Te KIRC CNV data of 10 CRGs was acquired and visualized from the cBioPortal database (https://www.cbioportal.org/, accessed on May 20, 2022) [37,38].

Immune Cell Infltration and Immune Checkpoint
Analysis. To assess the reliability results of the immune score evaluation, we used the "immundeconv" R package and the EPIC algorithm for further analysis [39]. Te algorithms had been benchmarked and had a unique advantage [40].

Western Blotting
Analysis. Ice-cold radioimmunoprecipitation assay bufer (Sigma-Aldrich, R0278, USA) was used to lyse cells and total proteins were extracted. We quantifed protein using BCA kit (Termo, 23227, USA) according to the instruction. Te same amount of protein was loaded for each sample and was transferred to PVDF membranes. Primary antibodies were incubated at 4°C overnight (β-actin, # 4970, Cell Signaling Technology; Anti-FDX1, ab108257, Abcam). Te secondary antibody (Anti-rabbit IgG, # 7074, Cell Signaling Technology) was added and incubated at room temperature for 1 h. Signals were detected by chemiluminescence (ECL Western Blotting Detection Reagents, GE Healthcare) and visualized using G: BOX Chemi Gel Documentation System (Syngene, Frederick, MD, USA).

Statistical Analysis.
Results were reported as mean ± SD for triplicate experiments unless otherwise indicated. Te statistical diferences of two groups were compared through the Wilcox test, and the signifcance diference of three groups was tested with the Kruskal-Wallis test. All analytical methods above and R packages were performed using R software version v4.0.3 (https://www.r-project.org) or GraphPad Prism v8.0.1. A value of p < 0.05 was considered as statistically signifcant, unless stated otherwise.

Te Potential Prognosis
Role of 10 CRGs. First, we investigated the expression level of 10 CRGs in ccRCC. Te results showed that only CDKN2A was upregulated and the

Construction and Validation of CRGs Prognostic Model.
Te LASSO Cox regression algorithm was used to construct the prognosis model in the training cohort for predicting clinical outcomes of ccRCC with CRGs. FDX1, PDHB, and CDKN2A were included to construct the risk model. Te risk score was   ). To further validate this prognostic model, the risk score was determined in the validation cohort, and the patients were categorized into two risk groups based on the cut-of value ( Figure 4(d)). Also, the high-risk group had a signifcant worse OS (Figure 4(e)). Te predicted AUCs of the ROC curve for 1-year, 3-year, and 5-year overall survival rates were 0.709, 0.582, and 0.614 (Figure 4(f)).

Construction and Evaluation of Nomogram.
Te age, pathologic T, pathologic M, pathologic stage, histopathological grade, and risk score could signifcantly afect the OS according to the univariate Cox analysis ( Figure 5(a)). Te multivariate Cox analysis showed that age, pathologic M, and risk score were independent risk factors for ccRCC patients in the training cohort ( Figure 5(b)). Furthermore, an easy-to-use and clinically adaptable nomogram was constructed. Te patients with higher total points were associated with worse 1-year, 3-year, and 5-year OS ( Figure 5(c)). Te calibration curve indicated the accuracy of the nomogram ( Figure 5(d)).

Lower FDX1 Expression Showed Worse Clinicopathologic
Features. To further explore the infuence of the three CRGs on ccRCC, univariate and multivariate Cox regression analyses were performed, and the results showed that only FDX1 was an independent protective factor in ccRCC (HR 0.525, 95% CI 0.338-0.816, p � 0.004, Table 2). Due to the critical role of FDX1 in cuproptosis, we focused on the FDX1 gene. Te expression level of FDX1 in diferent cancers was explored using the GEPIA2 database. Among the 33 cancer types in TCGA, the expression level of FDX1 was signifcant    0 1 2 3 4 5 6 7 8 9 10 11 12   0 1 2 3 4 5 6 7 8 9 10 11     Supplementary Figure S3B). The association between the expression of FDX1 and clinicopathologic features in ccRCC was analyzed. The results showed that lower expression of FDX1 was associated with males, a higher pathologic N stage, a higher pathologic T stage, a higher pathologic M stage, a higher pathologic stage, and a higher histologic grade ( Figure 6(e), Supplementary  Figures 3C and 3D).

Identifcation of the FDX1 Expression in the External
Dataset and Our Database. Te expression level of FDX1 was further identifed by the GEO database in the GSE53757 and GSE66271 databases, which contained 72 and 13 paired tumor tissues and adjacent normal tissues, respectively. FDX1 was signifcantly downregulated in tumor tissues (Figures 7(a) and 7(b)). We also explored the expression of FDX1 at the single-cell RNA-seq level in the GSE73121 database, which comprised 43 PDX primary ccRCC cells and 36 PDX metastatic ccRCC cells. FDX1 was lower expressed in metastatic ccRCC cells (Figure 7(c)). Also, we found that the single-cell RNA-seq results in the GSE159115 database demonstrated that the expression of FDX1 was lower in ccRCC tumor epithelial cells than proximal tubule cells, and the bulk-RNA seq in this study also identifed this result (Figure 7(d)). Te expression of the FDX1 protein was downregulated in tumor tissues when compared with normal tissues in the Human Protein Atlas (HPA) database (Figure 7(e)). And FDX1 was mainly expressed in mitochondria, according to immunofuorescence results in U2OS and A431 cells from the HPA database (Figure 7

Te Expression of FDX1 Was Correlated with DNA Methylation Level.
To further explore the underlying mechanism of the lower expression level of FDX1, we frst investigated the mutation status of 10 CRGs in TCGA-KIRC. Te results showed that only 2 (0.4%) of the ccRCC patients harbored copy number alterations (Supplement Figure S4A). We speculated that the low expression level of FDX1 may relate to DNA methylation. Ten, we compared the methylation levels of 11 CpG sites in FDX1 between 323 ccRCC tissues and 160 adjacent normal tissues (Figure 8(a)). Te detailed information of 11 CpG sites is shown in Table 3. Te results also indicated that 6 CpG sites were hypermethylated and 5 CpG sites were hypomethylated in ccRCC tissues when compared with adjacent normal tissues (Figure 8(b) and Supplementary Figure S4B). Te correlation between FDX1 expression and the methylation level of 6 CpG sites was analyzed. Linear correlation analysis results showed that only the methylation levels of the cg26061355 site were signifcantly negatively correlated with In order to better investigate the mechanism of FDX1 during ccRCC tumorigenesis, we compared the 75% high FDX1 expression group with the 25% low FDX1 expression group to fnd diferentially expressed genes. Te results showed that there were 251 genes upregulated in the high expression group and 164 genes downregulated in the high expression group (Supplementary Figure S5A and Figure S5B). GO and KEGG functional analyses were performed to explore the afected pathway that FDX1 regulated. Te results showed that fatty acid metabolism, glycolysis/ gluconeogenesis, pyruvate metabolism, oxidative phosphorylation, and citrate cycle (TCA cycle) pathways were upregulated (Figure 8(f )), and cytokine-cytokine receptor interaction, NF-kappa B signaling, TNF signaling, IL-17 Journal of Oncology 9 signaling, and mineral absorption pathways were downregulated in high expression groups (Figure 8(g), Supplementary Figure S5C-5F).

TMB and Immune Cell Infltration Analysis.
Te differences in mutation landscapes were found between FDX1 high and low expression groups (Figure 9(a)). In the  BRCA  CESA  CHOL  COAD  DLBC  ESBC  GBM  HNSC  KICH  KIRC  KIRP  LAML  LGG  LIHC  LUAD  LUSC  MESO  OV  PAAD  PCPG  PRAD  READ  SARC  SKCM  STAD  TGCT  THCA  THYM  UCEC  USC  high FDX1 expression group, the mutation rates were lower when compared with the low FDX1 expression group (74.27% vs. 88.20%). Moreover, the TMB between the FDX1 high expression group and the FDX1 low expression group was investigated, and the result demonstrated the TMB was lower in the high expression group (Figure 9(b)). We also found that FDX1 was downregulated in tumors with VHL, PBRM1, SETD2, BAP1, and KDM5C mutations (Supplementary Figures S6A-S6E). Te TMB was negatively correlated with FDX1 expression (Figure 9(c)). From the previous analysis of KEGG, we found that lower FDX1 could relate to the immune microenvironment. Ten, we compared immune cell infltration between the FDX1 high expression group and low expression group, which demonstrated that B cells, macrophages, and NK cells were higher in the FDX1 low expression group. T cell CD4+ and T cell CD8+ were lower in the FDX1 low expression group (Figure 9(d)). Besides, we detected a correlation between FDX1 and diferent immune cells. Results revealed that the expression level of FDX1 was positively correlated with T cell CD4+ infltration level and negatively associated with the infltration levels of B cells, macrophages, and NK cells (Figure 9(e), Supplementary Figures S6F-S6G).

Te Prediction of Immune and Target Terapy Responses.
Te expression of immune checkpoint genes was detected in diferent FDX1 expression groups, and the results showed that the expression levels of CD274 were signifcantly higher and those of CTLA4, LAG3, PDCD1, PDCD1LG1, and TIGIT were signifcantly lower in the FDX1 high expression group (Figure 10(a)). Moreover, the potential immune response was predicted. Te results demonstrated that immune response scores were higher in the lower expression group (Figure 10(b)). Due to the fact that FDX1 is a direct target of elesclomol, we also found that the patients in the FDX1 high expression group were signifcantly more sensitive to elesclomol, and the patients in the FDX1 low expression group were signifcantly more sensitive to sunitinib and axitinib (Figures 10(c)-10(e), Supplementary Figures S6H-S6J).

Disscussion
With the development of diagnostic technologies and immune checkpoint inhibitors, the survival of advanced ccRCC patients has been dramatically improved [43]. However, there were still some treatments that failed in patients with ccRCC because of metastasis, which could be attributed to the heterogeneity of ccRCC tissues [44,45]. Cuproptosis, a novel manner of programmed cell death, is diferent from ferroptosis, necrosis, apoptosis, and pyroptosis. Recently, a study revealed that cuproptosis was closely associated with mitochondrial respiration. However, ccRCC is mainly dependent on the glycolytic pathway for energy because of the loss of function of VHL and the accumulation of HIF-1α [22,46]. Tumor cells may develop the mechanism of cuproptosis tolerance to attenuate cell death and promote proliferation or metastasis. Te role of cuproptosis in patients with ccRCC has rarely been studied. In this study, we constructed a CRGs prognosis model and further elucidated the vital role of the key gene of cuproptosis, FDX1, in the ccRCC. Our study about CRGs genes may shed new light on tumor classifcation and response to treatment.
In the present study, we screened 3 diferentially expressed CRGs, FDX1, PDHB, and CDKN2A. All of them were associated with OS or PFS, and then they were determined to construct a CRGs prognosis model using LASSO regression analysis. Te model could divide KIRC patients into high-risk and low-risk groups. Te model demonstrated great survival prediction efciency and was validated in the training and validation cohorts. Also, the ROC curve of the model in the training and validation cohorts showed moderate diagnostic performance in predicting 1-year survival (0.675 and 0.708) and 5-year survival (0.668 and 0.614). Te reason that the model did not show a high level of performance may be due to the limited number of CRGs and the infuence of various pathways or cell death methods. Te other important pathway genes could be incorporated into this model in the future. To be more accessible for the model, we constructed a nomogram prediction model throughout, combining it with the independent clinicopathological risk factors. Ten, the calibration curves showed the high accuracy of the nomogram model. In general, we constructed a novel CRGs prognosis model, which could guide clinical surveillance and treatment decisions.
To further explore the prognostic role of FDX1, PDHB, and CDKN2A, we performed multivariate Cox regression to identify the key gene, and the results showed that low expression of FDX1 was an independent risk factor. Most importantly, FDX1 was a key gene that regulated cuproptosis. FDX1 has also been closely related to lipid-related and steroid metabolism [47,48]. FDX1 is essential for the synthesis of various steroid hormones (pregnenolone, aldosterone, and cortisol), and lower FDX1 expression is associated with increased glycolysis [49,50]. FDX1 promoted cuproptosis through increased protein lipoylation, and the deletion of FDX1 conferred resistance to copperinduced cell death [51]. Te role of FDX1 in cancer is rarely studied. Recently, a study showed that FDX1 was decreased, and the patients with lower expression of FDX1 had a worse prognosis for lung cancer [49]. Te role of FDX1 remains unclear. Terefore, we focused on the role of FDX1 in ccRCC.
Among the expression of FDX1 in 33 cancer types, KIRC was the only one cancer that was signifcantly downregulated. Lower FDX1 expression was associated with worse OS and PFS in KIRC, which showed that FDX1 played a vital role, especially in KIRC. Te results could be attributed to the tissue specifcity of FDX1 expression. In addition, the diferent FDX1 expression groups presented  a signifcant correlation with gender, pathological T stage, lymphatic invasion, metastasis, pathological stage, and histological grade in ccRCC. On the basis of these results, we concluded that FDX1 could be a valuable prognostic biomarker in ccRCC. Ten, the lower expression of FDX1 in tumor tissues was identifed in external and our own databases. Te external databases demonstrated that in the level of bulk-RNA seq or single-cell RNA seq, the FDX1 was downregulated in tumor tissues, metastasis tumor cells and ccRCC epithelial cells. Te HPA databases further confrmed this result, and the immunofuorescence analysis showed FDX1 was mainly distributed in mitochondria. In our own database, FDX1 was downregulated in renal cancer cell lines and paired tumor tissues at both of mRNA and protein levels. Overall, lower expression of FDX1 was associated with worse OS and PFS, especially in ccRCC, and may function as a tumor suppressor due to being downregulated in tumor cells. Te mechanisms of low expression of FDX1 were further explored in ccRCC. Te mutation status of CRGs was detected in the cBioPortal database, and the results showed low mutation rates and copy number alternation in the FDX1 gene. Ten, we detected that the methylation level was negatively correlated with the expression of FDX1 and verifed this in the GSE61441 dataset, which indicated that DNA hypermethylation might play a vital role in decreasing the expression of FDX1 in ccRCC. However, it was not strongly correlated between FDX1 expression and the methylation level of cg26061355, which indicated that there may be some other mechanisms contributing to the low expression of FDX1 in ccRCC. Previous studies reported that SF-1 and cJUN could bind with the promoter of FDX1 in MA-10 leydig cells and ovarian granulosa cells [52,53]. Tere were also some nucleotide polymorphisms in FDX1 that may contribute to IgA nephropathy [54,55]. All of these mechanisms may also contribute to the low expression of FDX1 in ccRCC. Furthermore, KEGG and GO analyses were performed to explore the underlying mechanisms of FDX1 in tumorigenesis. In the high FDX1 expression group, oxidative phosphorylation, the citrate cycle (TCA) cycle, and the gluconeogenesis pathway were upregulated, which indicated the correlation between high FDX1 and upregulated mitochondrial respiratory. Meanwhile, lower expression of FDX1 may be related to the IL-17 signaling pathway and T cell activation, which indicates that low expression of FDX1 may be related to immune cell infltration.
Te mutation landscapes were compared between two diferent expression groups. Te TMB were higher in the low expression group. Meanwhile, TMB was negatively correlated with the expression of FDX1. Ten, the immune cell infltration was analyzed, and the results showed that T cell CD4+ were lower expressed and positively correlated with the low FDX1 expression group. Te B cell, macrophage, and NK cell were more highly expressed and negatively correlated with the low FDX1 expression group. Higher B cells, macrophages, and lower T cell CD4+ were associated with poorer OS and PFS [56][57][58][59]. However, though higher NK cells were related to better OS, the population of NK cells was small in ccRCC [60]. Te CD274 was lower in the low FDX1 expression group, as were the CTLA4, LAG3, PDCD1, PDCD1LG2, and TIGIT, which were higher in the low FDX1 expression group. Te results helped us better choose ICB for immune treatment or provided a theoretical basis for the development of new therapeutic strategies based on immune characteristics [61,62]. Te immune response prediction results showed that low FDX1 expression had worse response. Tis may be due to low T cell CD4+ and T cell 8+ infltration in ccRCC. Tis result seems to contradict the higher TMB in FDX1 low expression groups [63,64]. However, a recent study showed that the TMB may not have a tight correlation with ccRCC [65]. In drug sensitivity analysis, elesclomol showed a better therapeutic efect in the FDX1 high expression group, while sunitinib and axitinib showed a worse therapeutic efect in the FDX1 low expression group. Also, as for the FDX1 high expression group, some new treatment strategies, such as drug delivery systems in combination with elesclomol, may have had a better performance in tumor treatment [66,67]. Above all, these fndings implied that the expression of FDX1 could predict immune and target treatment responses in ccRCC. (c-e) Sensitivity analysis for elesclomol, sunitinib, and axitinib in FDX1 high expression groups and low expression groups. Te statistical diferences between diferent groups were compared through the Wilcoxon test. * p < 0.05, * * p < 0.01, and * * * p < 0.001.
Despite the strengths of our study, some limitations should be acknowledged. A larger sample size would be required for further validation of CRGs' prognostic model. In addition, more experiments are required to verify the DNA methylation and immune correlates for the hub gene of cuproptosis, FDX1.
On the whole, we constructed CRGs prognostic model and observed excellent performance. Ten, we identifed the hub gene for cuproptosis. FDX1 was an independent prognostic biomarker, especially in ccRCC. Te lower expression of FDX1 may be related to DNA hypermethylation. FDX1 could be a biomarker, which could predict immune and target therapeutic responses. In conclusion, our study will contribute new insights to clinical surveillance and help physicians with treatment decision-making.

Data Availability
Te data supporting the results of this study are included in the article and can be consulted with the corresponding author on reasonable request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Authors' Contributions
KG and YG conceived the study and participated in its design. KZ, WY, ZZ, and KM performed the data acquisition and statistical analyses. KZ drafted the frst manuscript of the study. LL and YX wrote the manuscript. KZ, JQ, CY, LC, and JZ dealt with the fnal typesetting and revised the manuscript. All authors contributed to the manuscript revision and approved the submitted version. Kenan Zhang and Wuping Yang have contributed equally to this work and share frst authorship.

Supplementary Materials
Supplementary