Investigating the Role of FoxP3 in Renal Cell Carcinoma Metastasis with BAP1 or SEDT2 Mutation

Forkhead box protein P3 (FoxP3) primarily functions as the master regulator in regulatory T cells (Tregs) differentiation, but its high level of expression has also been found in tumor cells recently. The aim of our study was to clarify the role of FoxP3 in renal cell carcinoma (RCC) progression and metastasis. We verified the FoxP3 characteristic clinicopathological data from The Cancer Genome Atlas (TCGA) database using bioinformatics tools. Meanwhile, RNA sequencing was performed to determine the FoxP3 biofunction in RCC progression. Our results showed that high expression of FoxP3 was found in BAP1- or SETD2-mutant patients with RCC, and a higher FoxP3 expression was related to worse prognosis. However, there was no statistically significant relationship between the FoxP3 IHC score and RCC malignant progression owning to the limited number of patients in our tissue microarray. Using in vitro FoxP3 loss-of-function assays, we verified that silencing FoxP3 in 786-O and ACHN cells could inhibit the cell migration/invasion capability, which was consistent with the data from RNA sequencing in 786-O cells and from the TCGA datasets. Using an in vivo nude mice orthotopic kidney cancer model, we found that silencing FoxP3 could inhibit tumor growth. In conclusion, our study demonstrated that BAP1 or SEDT2 mutation could lead to higher expression of FoxP3 in RCC patients, and FoxP3 could eventually stimulate RCC cells’ invasion and metastasis, which might indicate that FoxP3 could function as a potential oncogene in RCC progression.


Introduction
Kidney cancer is the 14th most common global malignancy, with an estimated 431,288 new cases in 2020 [1]. Renal cell carcinoma (RCC) is the most common type of kidney cancer, accounting for nearly 3% of all cancers worldwide [2]. In recent years, the incidence of RCC has rapidly increased worldwide, including in China [3]. According to the World Health Organization's report, there were approximately 79,000 new cases of kidney and renal pelvis cancer in the United States in 2022. It has been found that the incidence of RCC increases exponentially with age and is approximately twice as common in men as in women [4]. Generally, clear-cell renal cell carcinoma (ccRCC) accounts for approximately 75% of RCC cases [5], and papillary RCC is the second most common tumor [6]. RCC is known to be radio-resistant and chemo-resistant, and the outcome of metastatic RCC is very poor [5,7,8]. Up to 17% of patients have distant metastases at the time of diagnosis [9,10]. In 2020, 179,368 people died from RCC globally [1]. For patients with stage III RCC, the five-year survival rate is approximately 60%; furthermore, the five-year survival rate for patients with stage IV RCC is less than 10% [6,11]. In recent years, some specific genes have been demonstrated to be associated with the development and progression of RCC. Chromosomal genetic disorders, such as von Hippel Lindau (VHL) syndrome with In this study, we performed gain-and loss-of-function assays of FoxP3 to explore its biofunction in RCC progression. Additionally, we performed immunohistochemistry assays to confirm whether FoxP3 is associated with the clinicopathological features of RCC and designed a nude mice orthotopic RCC model to verify FoxP3 function in vivo. Our preliminary data might suggest that FoxP3 could modulate the RCC tumor microenvironment and promote RCC aggressiveness.

FoxP3 Was Oncogenic and Correlated with Worse Clinical Outcomes
To explore and confirm the transcription level of FoxP3 in ccRCC, we analyzed the FoxP3 mRNA level in 23 different cancer types. As shown in Figure 1A, the transcription level of FoxP3 was compared with adjacent normal tissues from cancer patients, including ccRCC. Through analyzing the TCGA transcriptome and mutational data of ccRCC patients, we found that the FoxP3 mRNA level was significantly and positively correlated with mutations in the SET domain containing 2 (SETD2; p < 0.0001) and BRCA1 associated protein 1 (BAP1; p < 0.001) genes ( Figure 1B). We predicted that FoxP3 was dysregulated in ccRCC patients and might be associated with the clinicopathological factors and prognosis.
To investigate the role of FoxP3 in ccRCC clinical outcomes, the patients' overall survival (OS) was obtained in UALCAN for analyzing cancer OMICS data. As shown in Figure 1C, a higher FoxP3 expression was related to a worse outcome. Subsequently, we found that FoxP3 expression was associated with a high stage ( Figure 1D,E). We verified the SETD2 and BAP1 mutations and the FoxP3 expression in ccRCC cell lines. Consistent with the results showed in Figure 1B, the FoxP3 expression level was much higher in the BAP1 mutation (769-P and UM-RC-6) and SETD2 mutation RCC cell lines (A704, Caki-1, and A498) than their specific wild-type cell lines (OS-RC-2, Caki-2, and 786-O) ( Figure 1F). A higher FoxP3 expression level in tumor tissues than in benign tissues was also validated in the GSE781 datasets, which included nine tumor and eight benign tissues ( Figure 1G).
To further validate the relationship between FoxP3 expression level and the patients' clinicopathological characteristics, we performed FoxP3 immunohistochemistry (IHC) in a ccRCC tissue microarray, including 90 tumor tissues and their paired 90 adjacent normal kidney tissues. The IHC results demonstrated that the FoxP3 in tumor cell protein expression was higher in tumors than in their adjacent normal kidney tissues (Figure 2A), and a higher expression level of FoxP3 protein in tumor cells was related to higher TNM stages ( Figure 2B-D). However, the expression level of FoxP3 in tumor cells had no significant differences between different tumor grades in our 90 samples. Collectively, these data suggested that the expression level of FoxP3 was higher in tumor tissues, and that it might be positivity correlated with the ccRCC stages.

FoxP3 Could Facilitate RCC Tumor Metastasis
To examine the effect of FoxP3 on the biological processes in the entire network of genes, we performed high-throughput RNA sequencing to systematically analyze the changes in gene expression between the FoxP3-silenced 786-O cells and the controls. FDR < 0.05 and |log 2 FC| > 1 were screened out as significantly different genes and submitted to GSEA to run the hallmark gene sets. As shown in Figure 3A, there were "bumpy" enrichment plot gene sets in the control cells compared to the FoxP3-silenced 786-O cells ( Figure 3A). Furthermore, we found that the gene sets "hallmark_myc_targets", "hall-mark_mTOR1_targets", "hallmark_hypoxia_targets", "hallmark_wnt_targets", and "hall-mark_E2F_targets" were responsible for 786-O control cells for the biological behavior of ccRCC. We also submitted TCGA RNA-seq data to GSEA to run the hallmark gene sets, which included two groups, FoxP3 high-expression group (top 50) and FoxP3 lowexpression group (top 50) ( Figure 3B). The results showed that the curve top of "hall-mark_epithelia_mensenchymal_transition_targets" existed in the FoxP3 high-expression group. Thus, higher FoxP3 might be related to epithelial-mesenchymal transition (EMT) in ccRCC progression. CESC, cervical squamous cell carcinoma and endocervical adenocarcinoma; CHOL, cholangiocarcinoma; COAD, cholangiocarcinoma; ESCA, esophageal carcinoma; GBM, glioblastoma multiforme; HNSC, head and neck squamous cell carcinoma; KICH, kidney chromophobe; KIRC, kidney renal clear cell carcinoma; KIRP, kidney renal papillary cell carcinoma; LIHC, liver hepatocellular carcinoma; LUAD, lung adenocarcinoma; LUSC, lung squamous cell carcinoma; PAAD, pancreatic adenocarcinoma; PRAD, prostate adenocarcinoma; PCPG, pheochromocytoma and paraganglioma; READ, rectum adenocarcinoma; SARC, sarcoma; SKCM, skin cutaneous melanoma; THCA, thyroid carcinoma; THYM, thymoma; STAD, stomach adenocarcinoma; UCEC, uterine corpus endometrial carcinoma. (B) Distribution of FoxP3 expression level in ccRCC with mutations in VHL, PBRM1, BAP1, and SETD2. The blue part shows the expression level of FoxP3 ranked from high to low. The distribution below shows whether the four genes are mutated at the corresponding FoxP3 expression levels. The red vertical line indicates that the gene was mutated, and the grey vertical line indicates that the gene was not mutated. The p-values on the right side indicate the  , and T3 (6) of ccRCC in the tissue microarray, respectively. IHC staining was graded as follows: 0 for 0%, 1 for ≤25%, 2 for 25-50%, 3 for 50-75%, and 4 for ≥75%.
The IHC intensity was scored as follows: 0 for no staining, 1 for weakly positive staining, 2 for moderately positive staining, and 3 for strongly positive staining (ns: no statistical significance).

FoxP3 Could Facilitate RCC Tumor Metastasis
To examine the effect of FoxP3 on the biological processes in the entire network of genes, we performed high-throughput RNA sequencing to systematically analyze the changes in gene expression between the FoxP3-silenced 786-O cells and the controls. FDR < 0.05 and |log2FC| > 1 were screened out as significantly different genes and sub-  (23), and T3 (6) of ccRCC in the tissue microarray, respectively. IHC staining was graded as follows: 0 for 0%, 1 for ≤25%, 2 for 25-50%, 3 for 50-75%, and 4 for ≥75%.
The IHC intensity was scored as follows: 0 for no staining, 1 for weakly positive staining, 2 for moderately positive staining, and 3 for strongly positive staining (ns: no statistical significance).  the difference in expression of the "Hall-mark_epithelia_mensenchymal_transition_targets" pathway between the FoxP3 high-expression group and the FoxP3 low-expression group. The top part of the graph was described previously. The red part in the middle of the graph indicates that the gene was highly expressed in the FoxP3 high-expression group. The blue part indicates that the gene was highly expressed in the FoxP3 between the FoxP3 high-expression group and the FoxP3 low-expression group. The top part of the graph was described previously. The red part in the middle of the graph indicates that the gene was highly expressed in the FoxP3 high-expression group. The blue part indicates that the gene was highly expressed in the FoxP3 low-expression group. The lower part of the graph shows the rank in the ordered dataset, where a value greater than 0 means that the gene was more highly expressed in the FoxP3 high-expression group. The lower the vertical line is above 0, the higher the expression of the corresponding gene in the To verify the clinical sample bioinformatics analysis results, we performed FoxP3 loss-of-function assays in 786-O and ACHN cell lines. As shown in Figure 3C, the wound healing assay results demonstrated that the FoxP3 inactivation impaired the 786-O and ACHN cells' migration by 60-80% at 72 h. We also applied a Transwell assay to assess the impact of FoxP3 on the ccRCC migration capability ( Figure 3D). The results further demonstrated that silencing FoxP3 could significantly impair the 786-O and ACHN cells' migration (786-O, p = 0.03; ACHN, p = 0.01) and invasion capabilities (786-O, p = 0.01; ACHN, p < 0.01). Our findings verified that a high expression of FoxP3 could promote ccRCC cells' migration and invasion in vitro.

FoxP3 Could Promote RCC Tumor Immune Evasion
The immune function of FoxP3 in Treg cells is clear, but the immune function of FoxP3 in cancer is unclear, especially in RCC progression. To understand whether the FoxP3 expression in tumor cells affects the RCC microenvironment, we analyzed the highthroughput RNA sequencing data in 786-O cells compared with silencing FoxP3 in 786-O cells. The GSEA hall_mark_gene set results showed the immune-related signaling pathways "hallmark_IL2_stat5_signaling", "hallmark_THFA_signaling_via_NFKB_targets", "hall-mark_inteferon_alpha_response_targets", "hallmark_interferon_gamma_repsonse_targets", "hallmark_inflammatory_response_targets", and "hallmark_IL-6_JAK_stat3_response_signaling" ( Figure 4A). In addition, the TCGA ccRCC data of the FoxP3 high-expression group (top 50) and the FoxP3 low-expression group (top 50) were sent to run hallmark gene sets. The results were consistent with our above data. "Hallmark_IL2_stat5_signaling", "hall-mark_inteferon_alpha_ response_targets", "hallmark_interferon_gamma_repsonse_targets", and "hallmark_ inflammatory_response_targets" were all enriched in the FoxP3 highexpression group ( Figure 4B). We also found that the patients in the FoxP3 high-expression group had more Tregs infiltration in the tumor microenvironment ( Figure 4C). Collectively, the results from our cell lines and TCGA ccRCC RNA-seq results indicated that FoxP3 might play a pro-oncogenic role in ccRCC by activating immune-related pathways and by recruiting more Treg cells into the tumor microenvironment. ol. Sci. 2023, 24, x FOR PEER REVIEW 9 of 18

FoxP3 Silencing Inhibited RCC Growth in Mouse Xenograft Tumors
For the FoxP3 loss-of-function in vivo experiment, the 786-O cell line was selected in our study, because (1) 786-O is established as one of the first ccRCC cell lines and it has many characteristics of ccRCC; meanwhile, ccRCC is the most dominated pathological type among all RCC cases; (2) compared to other RCC cell lines, there was a higher expression level of   The two groups had similar rates of increase in weight ( Figure 5B). However, the shNC group had a larger average tumor volume than the shFoxP3 group ( Figure 5C). Collectively, shFoxP3 might inhibit RCC tumor mass growth in vivo.

FoxP3 Silencing Inhibited RCC Growth in Mouse Xenograft Tumors
However, the shNC group had a larger average tumor volume than the shFoxP3 group ( Figure 5C). Collectively, shFoxP3 might inhibit RCC tumor mass growth in vivo. The distribution of the body weight change of the 2 groups of mice in the experiment within 30 days after the injection. The horizontal line represents the time recorded after injection, and the vertical line represents the body weight change in the two groups, respectively (error bars indicate ± SD, n = 6; ns: no statistical significance). (C) Images of the kidneys together with tumor masses of nude mice in the two groups after the experiment. Nude mice were euthanized 30 days after injection. One nude mouse in the FoxP3-silenced group was euthanized early at 24 days after injection due to poor health status. The red circled area shows the tumor masses growing from the subcapsule of the kidneys of the nude mice.

Discussion
It is well known that RCC accounts for 3% of all cancers worldwide [2], and the current prognosis of patients with intermediate to advanced RCC is not very favorable [6,11]. Within the treatment and management options for patients with metastatic RCC, molecular targeted drugs and immune check-point inhibitor therapies have been heavily investigated. Previous studies have shown that Tregs are a class of cells that can regulate the body's immunity [19] and may play different roles in tumors at different locations [31,[34][35][36]; their biological functions can be regulated by the expression of FoxP3 [15,16]. In recent years, it has been found that some normal tissues can also express FoxP3 in addition to immune cells [30][31][32][33]. Therefore, exploring the expression and the biofunction of FoxP3 in RCC tumor cells may help physicians to make better decisions in the treatment modalities. In our study, we constructed FoxP3 in vivo and in vitro experiments, and we demonstrated that the expression level of FoxP3 in ccRCC was increased by the induction of BAP1 and SETD2 mutation. The up-regulation of FoxP3 could facilitate the migration and invasion of ccRCC cell lines, activate the immune-suppression-related pathways, and attract more Tregs infiltrated into the tumor microenvironment. All these would promote cancer cells immune evasion and metastasis,and eventually it could lead to worse outcomes for patients with ccRCC. Nude mice were euthanized 30 days after injection. One nude mouse in the FoxP3-silenced group was euthanized early at 24 days after injection due to poor health status. The red circled area shows the tumor masses growing from the subcapsule of the kidneys of the nude mice.

Discussion
It is well known that RCC accounts for 3% of all cancers worldwide [2], and the current prognosis of patients with intermediate to advanced RCC is not very favorable [6,11]. Within the treatment and management options for patients with metastatic RCC, molecular targeted drugs and immune check-point inhibitor therapies have been heavily investigated. Previous studies have shown that Tregs are a class of cells that can regulate the body's immunity [19] and may play different roles in tumors at different locations [31,[34][35][36]; their biological functions can be regulated by the expression of FoxP3 [15,16]. In recent years, it has been found that some normal tissues can also express FoxP3 in addition to immune cells [30][31][32][33]. Therefore, exploring the expression and the biofunction of FoxP3 in RCC tumor cells may help physicians to make better decisions in the treatment modalities. In our study, we constructed FoxP3 in vivo and in vitro experiments, and we demonstrated that the expression level of FoxP3 in ccRCC was increased by the induction of BAP1 and SETD2 mutation. The up-regulation of FoxP3 could facilitate the migration and invasion of ccRCC cell lines, activate the immune-suppression-related pathways, and attract more Tregs infiltrated into the tumor microenvironment. All these would promote cancer cells immune evasion and metastasis, and eventually it could lead to worse outcomes for patients with ccRCC.
FoxP3 is a key transcription factor in Tregs development and function and has been extensively studied [25,29]. In Tregs, the FoxP3 promoter is regulated by forkhead box O protein 1 (FOXO1), forkhead box O protein 3 (FOXO3), and other regulatory elements [38,39]. FoxP3 transcription is highly controlled by conserved noncoding sequences (CNSs), which can interact with some important transcription factors, such as human mothers against decapentaplegic homolog 3 (SMAD3), avian erythroblastosis virus E26 oncogene homolog 1(ETS1), RUNX1, REL, etc. [38,40]. However, the regulation of FoxP3 in tumor cells is largely unclear, especially in RCC. In our study, we analyzed the FoxP3 transcriptome across TCGA cancers, and we found that the FoxP3 was highly expressed in the tumor tissues of ccRCC patients with BAP1or SETD2-mutant genotype ( Figure 1B). Then, we conducted an analysis in the CCLE database and found that the FoxP3 expression was higher in RCC cell lines with BAP1 or SETD2 mutation than their wild-type cell lines ( Figure 1F). In addition, we analyzed the staging of ccRCC with FoxP3 in the GEO database and performed IHC in tissue microarrays from patients with ccRCC. The data showed that the FoxP3 expression level was higher in tumors than in normal kidney tissues (Figures 1G and 2A). This was similar to the study by Sell K et al., who performed RT-PCR on kidney tissue samples from RCC patients and found that the FoxP3 expression levels in tumors were higher than those of adjacent normal kidney tissues [37]. However, unlike Sell K et al., who mainly focused on the aggregation of Tregs, our results might be associated not only with excessive Tregs but also with the overexpression of FoxP3 in ccRCC tumor tissues caused by BAP1or SETD2-mutant, and our findings might not have been reported yet. We also found that ccRCC tumors with a high stage could express higher levels of FoxP3 than those with a low stage ( Figure 2B-D), although no significant differences were found between different tumor grades. In general, we were able to observe that the high levels of FoxP3 were often associated with poor tumor staging for patients with ccRCC. Interestingly, Hakimi AA et al. analyzed 188 ccRCC patients for genetic sequencing and prognosis and found that BAP1 and SETD2 mutations were associated with worse cancer-specific survival (CSS) [41]. A study of over 1000 ccRCC patients reached similar conclusions by Manley BJ et al., who found that ccRCC patients with mutations in BAP1 and SETD2 were associated with short CSS and recurrence-free survival, respectively [42]. These studies further validated our results that the poor prognostic outcome associated with mutations in BAP1 and SETD2 was likely to be related to the high expression level of FoxP3. In recent years, it has been found that the oncological behavior from the two genotypic mutations might be different, with previous studies suggesting that SETD2 was associated with distant metastasis in ccRCC [43]. However, Peña-Llopis S et al. found that the BAP1 mutation might promote tumor cells' growth and make the prognosis of ccRCC patients even worse [44]. Although both gene mutation types resulted in the increased FoxP3 expression in RCC, their different oncological behaviors still needed to be specifically analyzed by physicians when appropriate treatment strategies were considered. In summary, FoxP3 was highly expressed in ccRCC with BAP1or SETD2-mutant with advanced staging and could lead to poor prognosis for patients with ccRCC.
Most of the current studies are mainly focused on the effect of FoxP3 in Treg cells in RCC. However, in our study, we also found that high expression of FoxP3 in ccRCC cells was associated with patients' poor prognosis. In gastric cancer, FoxP3 could promote gastric cancer migration and invasion through the TGF-β pathway [35]. In breast cancer, FoxP3 could induce the transcriptional activity of miR-200c and miR-141, which were elevated in patients with metastatic breast cancer [45]. In non-small cell lung cancer, FoxP3 could promote tumor metastasis through the Wnt/β-catenin signaling pathway and EMT [36]. However, no report has described the FoxP3 function in RCC progression. To explore the biological functions of FoxP3, we constructed the 786-O cells RNA sequencing experiment and analyzed the transcriptomic data. The GSEA showed that several gene sets of EMT-related genes were regulated by FoxP3 (Figure 3A), and the "hallmark_epithelial_mesenchymal_transition" had a "bumpy" enrichment in the FoxP3 high-expression group in the TCGA datasets ( Figure 3B). Our wound healing assay and Transwell assay showed that FoxP3 could promote RCC cells' migration and invasion ( Figure 3C,D). Consistent with our findings, several investigators have found similar results in different types of tumors. Previously, we mentioned that FoxP3 has been shown to promote the EMT pathway in non-small cell lung cancer [36]. Wang L et al. found that increased FoxP3 could lead to increased expression levels of miR-664a-3p, which might activate the EMT pathway and promote tumor progression in gastric cancer [46]. In addition, it has also been shown that FoxP3 regulates the expression of LINC00885, and that high expression of FoxP3 could promote cervical cancer cells' proliferation and the activation of EMT pathways [47]. These results suggest that FoxP3 could regulate different pathways of the EMT and promote cancer metastasis. Our study not only confirmed that FoxP3 could activate the EMT pathway in ccRCC, but also provided related genes' alteration by the high expression of FoxP3. Our results might indicate that we need to pay more attention to the high expression of FoxP3 in RCC, because it could be one of the main factors leading to RCC metastasis.
Previous studies have demonstrated that increased CD4 + CD25 + FoxP3 + Treg cells in RCC were related to worse outcomes [33], and the FoxP3 + tumor cells have been detected in the tumor-normal tissue borders [37]. In our study, FoxP3 could facilitate a tumor immune-suppressed microenvironment showed by 786-O RNA-sequence data and TCGA data ( Figure 4A,B), and we found that more Treg cells could infiltrate into tumor tissues with an abundant expression of FoxP3 ( Figure 4C). However, in the study by Chakiryan et al., they analyzed the association between common somatic mutations in ccRCC and the tumor microenvironment, and they found that SETD2 mutations were associated with significantly reduced levels of FoxP3 + T cells in tumors, stroma, and the tumor-stroma interface [48]. It has also been shown that the SETD2-mediated methylation pathway could inhibit IFN-α/β receptor signaling, which might ultimately impair the function of Tregs [49]. This result does not contradict our study; it might suggest that the SETD2 mutations have a reduced ability to convene Tregs compared to other mutant phenotypes, and the results still need to be further explored. Our results were consistent with Hiroyuki et al.'s results, who demonstrated the FoxP3 expression in non-small-cell lung cancer cells with tumorinfiltrating Tregs [50]. In pancreatic ductal adenocarcinoma, the FoxP3 expression was demonstrated in pancreatic cancer cells with tumor-infiltrating Tregs through the FoxP3/C-C motif chemokine ligand 5 (CCL5)/C-C motif chemokine receptor 5 (CCR5) pathway [27]. In murine melanoma, it was demonstrated that FoxP3 shifted the environment toward an immunosuppressive response by modifying the immune system [51]. These findings are quite similar to our results, suggesting that high expression of FoxP3 could activate immunerelated pathways, recruit more Treg cells' infiltration into the tumor microenvironment, and eventually promote immune evasion for cancer cells. In addition, our nude mouse orthotopic kidney cancer experiments showed that the growth of ccRCC cells with FoxP3 knocked out was significantly inhibited compared to controls ( Figure 5C). This result further suggests the role of FoxP3 for ccRCC growth in vivo.
There are very few studies on the effect of FoxP3 produced by RCC cells themselves on their growth, and more studies have explored the effect of Tregs on RCC. Tregs can suppress the body's immune response [19]; for example, Liotta F et al. found that Tregs exhibited inhibitory activity to effector T cells isolated from kidney tumors in vitro [52]. Ning et al. found that increased Tregs in tumor infiltration were positively correlated with VEGF protein expression [53]. In contrast, our study firstly proposed the relevant genes causing increased FoxP3 expression in RCC and identified possible pro-growth pathways and immune escape pathways induced by FoxP3 expression in RCC cells. These pathways demonstrated a novel role for FoxP3 in RCC growth, and might be used to further elucidate the effect of immunopharmaceutical treatment; they could also help in the development of other treatment regimens. The main limitations for our study were that most of our experiments were conducted in vitro, and that there was a lack of sufficient clinical data from RCC patients and a lack of study of detailed mechanisms to demonstrate that the BAP1or SETD2-mutant could indeed activate the specific growth, metastatic, and immune escape signaling pathways in RCC compared to other genotypic mutations. In the future, with the development of studies for novel genes, such as FoxP3, and their related regulatory mechanisms in RCC progression exploring a more precise and effective individualized therapy might be a possible direction.

Patient Samples and Bio-Information Analysis
FoxP3 transcriptome (count and FPKM value) across TCGA cancers (The Cancer Genome Atlas) data were analyzed on the website (http://gepia2.cancer-pku.cn/#index, accessed on 6 June 2021). The signature score was calculated by mean value of log2 (TPM + 1) of FoxP3 gene in gene set. The red box indicated the tumor samples while the blue one represented the normal tissues [54]. The effect of the FoxP3 expression on the overall survival (OS) and disease-free survival (DFS) was determined in UALCAN, which is an interactive web resource using TCGA datasets to analyze data from cancer patients. The mRNA expression of FoxP3 in ccRCC of different stages/grades was determined online using the UALCAN website (http://ualcan.path.uab.edu, accessed on 15 August 2021). Individual cancer stages were based on AJCC (American Joint Committee on Cancer) pathologic tumor stage information, and samples were divided into stage I, stage II, stage III, and stage IV groups. Tumor grade information was available in the TCGA database, and samples were categorized into grade 1, grade 2, grade 3, and grade 4 groups. Using the CPAN module "Statistics::Descriptive", mean TPM values (10 or above were retained) of each gene in normal samples and tumor samples were obtained separately [55]. The GSE781 dataset in the GEO database (Gene Expression Omnibus, https://www.ncbi.nlm. nih.gov/geo/, accessed on 18 October 2021) was accessed to examine FoxP3 expression in ccRCC tumor and normal tissues. The FoxP3 expression with relationships to PBRM1 MT , VHL MT , SETD2 MT , and BAP1 MT in ccRCC was analyzed using TCGA KIRC transcriptome (count and FPKM value) and mutational data. In the analysis, we first removed the lowvalue genes using a heterogeneity analysis; then, we normalized the data sets using the variance stabilizing transformation (VST) method in the DESeq2 package, as previously described [56].

CCLE Analysis in RCC Cell Lines
The FoxP3 RNA expression in RCC cell lines was determined by CCLE (Cancer Cell Line Encyclopedia (https://portals.broadinstitute.org/ccle/home, accessed on 10 November 2021), including 1000 cell lines' gene expression, DNA copy numbers, histone profiling, RNA-seq, and DNA methylation from more than 20 cancer types [57]. The gene mutation information of VHL, PBRM1, BAP1, and SETD2 in ccRCC cell lines was applied according to Wei X et al.'s article [7].

Gene Set Enrichment Analyses
We downloaded 611 ccRCC patients' RNA-seq data from the TCGA database. For the Gene Set Enrichment Analysis (GSEA), the FoxP3 high-expression group (top 135 (25% of 611)) and FoxP3 low-expression group (top 134 (25% of 611)) were set up in our study. FoxP3 Text.gct and FoxP3 text.cls were submitted to GSEA 4.1.0 version, and the hallmark gene sets were selected for the analysis.

Cell Culture
The 786-O and ACHN cells were purchased from American Type Culture Collection (ATCC, https://www.atcc.org, accessed on 5 December 2021) and authenticated by STR profiling in one year. The 786-O and ACHN cells were cultured with an RPMI medium (Cytiva, Shanghai, China) with 10% fetal bovine serum (Biological Industries, Kibbutz Beit Haemek, Israel) at 37 • C in a humidified 5% CO 2 incubator.

Wound Healing Assay
The wound healing assay was performed as previously described [59]. After 48 h of shFoxP3 plasmid transfection, 786-O and ACHN cells were seeded on 6-well plates and reached 100% confluence. The cells were starved overnight, and a wound was made by using a sterile 200 µL pipette tip to scratch the artificial wounds. The cells were washed with PBS 3 times. Wound healing was observed by microscopy after 24, 48, and 72 h.

Transwell Assay
The 786-O and ACHN cells were harvested after 48 h of shFoxP3 plasmid transfection. Then, 5 × 10 4 cells in 200 µL serum-free RPMI 1640 were added into the upper chambers containing 8 µM pore polycarbonate membrane filters (Millipore, Burlington, MA, USA). For the invasion assay, 5 × 10 4 cells in 200 µL serum-free RPMI 1640 were added into the upper chamber inserts with Matrigel (BD Biosciences, Franklin Lakes, NJ, USA), which had been plated 4 h in advance. Then, 800 µL RPMI 1640 medium containing 10% FBS was added to the lower chambers. After 24 h, the Transwell inserts were fixed with 4% paraformaldehyde for 10 min and stained with 0.1% crystal violet for 15 min at room temperature. Then, the migrating and invading cells were captured and counted under a light microscope.

Animal Experiments
Twelve BALB/c nude mice (6 weeks old, male) were randomly separated into two groups and injected with 2 × 10 6 786-O cells (control or shFoxP3 cells) into the subcapsule of both kidneys of each mouse. Our animal experiments were conducted under the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines and approved by the institutional review board of the First Affiliated Hospital of Xi'an Jiaotong University. The animals were euthanized on the 30th day or when the mice showed clear signs of ill health. The animal weights were measured every 3 days for 30 days.

Statistical Analysis
Each experiment was repeated 3 times. Differences between two groups (Student's t-test) were compared using the GraphPad Prism software (Version 6.0 software, GraphPad, Boston, MA, USA), and data are shown as the mean ± SD with error bars. p < 0.05 was considered statistically significant in our study.

Conclusions
Our study demonstrates that FoxP3 was increased in RCC cells with BAP1or SETD2mutant. Increased FoxP3 could not only activate the EMT pathway in RCC tumors, but also induce the immunosuppressive microenvironment of tumors and eventually attract more Tregs cells to activate RCC growth and metastasis. Approaches to investigating the specific mechanism of FoxP3 in RCC progression should be explored in the future [60].