Comprehensive Pan-Cancer Analysis of 33 Human Cancers Reveals the Immunotherapy Value of FOXP3

Background: Previous studies reported that FOXP3 is involved in the regulation of tumor microenvironment and the regulation of tumor local immunity. However, its mechanism has not been fully studied. This study aims to explore the potential relationship between FOXP3 and cancer immunotherapy in 33 human cancers. Method: The gene expression data and clinical characteristics of 33 cancers were retrieved from the Cancer Genome Atlas database. The immunotherapy cohort includes GSE157893, GSE67501, and IMvigor210, which come from a comprehensive gene expression database and are included in previously published studies. Analyze clinical parameters, including patient age, gender, and tumor stage to assess the prognostic value of FOXP3. At the same time, we conducted survival analysis and correlation analysis of tumor microenvironment. The correlation between FOXP3 and immunosuppressive agents and stimulants, as well as major histocompatibility complexes was also analyzed. Potential pathways related to FOXP3 signaling in cancer have also been explored. In addition, the correlation between FOXP3 and two immunotherapy biomarkers (tumor mutation burden and microsatellite instability) was studied. Finally, the immunotherapy response relationship between FOXP3 and the immunotherapy cohort was explored.


Introduction
Humoral immunity and cellular immunity are the two main effect mechanisms of human anti-tumor immunity. Of the two, cellular immunity plays a more important role in anti-tumor. Among them, T lymphocytes are the main participating cells. Regulatory T cells are the main cell population for antitumor immunosuppression. There is a xed ratio of immune cells in the normal body. If this ratio is changed, it will lead to the body's immune tolerance or weakened immune function, which will allow tumor cells to escape the host immune system. Identi cation and attack [1,2]. Studies have found that Treg plays an important role in tumor immunity. A high proportion of regulatory T cells have been found in various tumors such as non-small cell lung cancer and gastric cancer, and they are positively correlated with tumor metastasis and invasion [3][4][5]. Fork head box protein 3 (FOXP3) is an important member of the fork head/wing-shaped spiral transcription factor family and is mainly expressed in Treg. It is a landmark molecule of Treg. It regulates regulatory T cells (regulatory T cells). cells, Treg) and immune suppression [6,7]. At the same time, FOXP3 can exert an immunosuppressive function by regulating the content of anti-in ammatory factors such as IL-10 and TGFβ, and the expression of thousands of proteins in the body is also regulated by it [8]. TGF-β1 is a classic inhibitory cytokine that inhibits the immune function of DCs in the tumor microenvironment. Down-regulating the expression of FOXP3 can signi cantly reduce the secretion of TGF-β1. It can be seen that FOXP3 participates in the regulation of tumor microenvironment and the regulation of tumor local immunity. [9]. Studies have reported that the high expression of FOXP3 in tumor tissues due to the impaired tumor immune function leads to poor antitumor response. In breast cancer, cervical cancer, liver cancer, lung cancer and other tumor tissues, the increased expression of FOXP3 is associated with poor anti-tumor response, but the high expression of FOXP3 has been reported to be associated with good clinical results in patients with colorectal cancer [10][11][12]. However, few previous studies have paid attention to the immunotherapy value of FOXP3 in human cancer.
In this study, the expression of FOXP3 in 33 different cancers was demonstrated, and the potential impact of FOXP3 on the tumor immune microenvironment was studied. At the same time, in this article, we studied the relationship between FOXP3 expression and tumor mutational burden (TMB), microsatellite instability (MSI) and other immunomodulators and dynamic immune biomarkers. The relationship between FOXP3 expression and three immunotherapy cohorts was analyzed. In general, the research in this article provides a basis for the immunotherapy of FOXP3 in cancer, which may provide directions for further experiments.

Materials And Methods
Data collection: The expression data, mutation data, survival data and pan-cancer clinical data of 33 cancers are from UCSC Xena: (http://xena.ucsc.edu/). For the treatment cohort, a systematic search was conducted to determine the immune checkpoint lockout cohort, which can be retrieved and reported complete clinical information publicly. This study nally adopted three immunotherapy cohorts: engineered IL-15 cytokine mutein fused with anti-PD1 antibody can improve T cell function and antitumor immunity in tumors (GSE157893 cohort downloaded from GEO); use of Niwudan Treatmentresistant renal cell carcinoma (GSE67501 cohort downloaded from GEO); advanced urothelial carcinoma and atezolizumab intervention (IMvigor210 cohort downloaded from a previously published study) [13].
The clinical correlation between FOXP3 expression and various cancers Use the limma package in the R studio software to extract the expression level of the target gene FOXP3 and analyze the difference of the target gene FOXP3 to observe whether there is a difference between normal tissues and tumor tissues. We also investigated whether FOXP3 expression is different between different clinical groups (gender, age, and tumor stage). At the same time, use the limma package, GSVA package and GSEABase package in the R studio software to score gene activity.
According to the expression level of the target gene, it is divided into two groups of high and low expression. The survival package in R is used to perform Kaplan-Meier survival analysis to study the timedependent value of FOXP3 in 33 cancers. The contents of the study include overall survival (OS: the time from the start of treatment to death), disease-free survival (DFS: the period from the start of treatment to the recurrence of the disease), and disease-speci c survival (DSS: the outcome index changes to speci c Death caused by disease), progression-free survival (PFS: the period from the start of treatment to the progression of the disease). P value <0.05 is considered to be a difference in survival between the high and low expression groups.
Analyze the potential relationship between FOXP3 expression and immune-related factors First, use the ESTIMATE package in the R studio software to calculate the matrix score and immune score of each case; ESTIMATE is a tool used to predict tumor purity and the presence of in ltrating matrix and immune cells in tumor problems [14]. The ESTIMATE algorithm is based on a single-sample GSEA and generates four nal scores: matrix score (the higher the score, the more stromal cells in the tumor tissue), the immune score (the higher the score, the more immune cells in the tumor tissue), and the comprehensive score ( The higher the score, the higher the content of stromal cells and immune cells in the tumor tissue), and the purity of the tumor (the higher the overall score, the lower the purity of the tumor in the sample). Subsequently, use the ggplot2 package, ggpubr package and ggExtra package in the R studio software to analyze the correlation of the tumor microenvironment; use the e1071 package, preprocessCore package and Linna package in the R studio software to perform the immune cell in ltration analysis to obtain the immunity in each sample Cell content, followed by correlation analysis of immune cells to get the correlation between FOXP3 expression in tumor tissues and immune cell content; then correlation analysis of immune genes, in addition, through the TISIDB website (http://cis. hku.hk/TISIDB/index.php) explored the potential relationship between FOXP3 expression and immunosuppressive genes, immunostimulatory genes and MHC-related genes. Then highlight the two most relevant results and show them in the graph. Finally, in order to further study related signaling pathways, GSEA analysis was performed to determine the differential pathways between the low FOXP3 expression group and the high FOXP3 expression group. These pathways were from the Kyoto Encyclopedia of Genes and Genome database. If the relevant signal pathway meets a speci c standard (p<0.05), save the relevant signal pathway in the output graph. If the relevant signal pathway in the enrichment result is greater than 5, the top 5 most signi cant enrichment will be displayed. If the result of the collection pathway is less than 5, it shows all the enrichment pathways.

Immunotherapy response analysis
This study included and analyzed three related independent immunotherapy cohorts. Under normal circumstances, immunotherapy produces four outcomes: complete remission (CR), partial remission (PR), disease progression (PD), and stable disease (SD). Perform immunotherapy analysis on three data sets to study whether the expression of target gene FOXP3 affects the effect of patients receiving immunotherapy. In this study, those who achieved complete remission and partial remission were classi ed as the response group, and those who achieved disease progression and stable disease were classi ed as the non-response group, and then the two groups were compared.

Result FOXP3 expression in 33 cancers
In order to better show the method and content of this research, we show the details in Figure 1. The 33 cancers involved in this research are shown in Table 1. In Figure 2A, the expression of FOXP3 is different in 33 types of tumors (CESC, CHOL, COAD, HNSC, KIRC, LUAD, PRAD, SKCM, UCEC). In Figure 2B, FOXP3 is highly expressed in elderly patients in the ESCA, LAML, LGG, and LUAD groups, but is low in elderly patients in the BRCA and SKCM groups. Figure 2C shows that FOXP3 expression is different in different stages of certain tumors, including ACC, COAD, ESCA, KICH, KIRC, KIRP, SKCM, STAD, TGCT, and THCA. Figure 2D shows the difference in the expression of FOXP3 in different genders of certain tumors, including BLCA, KIRC, LUSC, and PAAD. In Figure 3, there is a signi cant positive correlation between the expression of FOXP3 in CESC, HNSC, OV, SKCM, and UCEC and overall survival. There is a negative correlation between the expression of FOXP3 in ACC, GBM, KIRC, and THYM and overall survival. Figure 4 shows that in UCEC, the expression of FOXP3 is positively correlated with disease-free survival, diseasespeci c survival, and progression-free survival. The expression of FOXP3 is positively correlated with disease-free survival in BLCA. In disease-speci c survival and progression-free survival, the expression of CESC, HNSC and FOXP3 are positively correlated, and GBM and KIRC are negatively correlated.
The relationship between FOXP3 expression and immune-related factors immunomodulators was also studied. In the analysis of immunosuppressive agents, Figure 7 shows that FOXP3 expression is positively correlated with BRCA TIGIT, and negatively correlated with THCA KDR. Figure 8 shows that in the analysis of immunostimulants, the expression of FOXP3 is positively correlated with the ICOS of ESCA and negatively correlated with the CD276 of TGCT. Figure 9 shows that FOXP3 expression is positively correlated with HLA-DRA of THCA, and negatively correlated with HLA-DMB of LGG. In Figure 10, considering the strong correlation between FOXP3 and THCA, BRCA, TGCT, ESCA, and LGG, we conducted a GSEA analysis to study the relationship between FOXP3 signaling pathways and their relationships. The pathway CELL ADHESION MOLECULES CAMS was enriched in the FOXP3 high expression group in ESCA and THCA. In Figure 11, FOXP3 expression is positively correlated with tumor mutation burden in THYM, UCEC, BRCA, and LGG, and negatively correlated with DLBC, HNSC, LUAD, LUSC, PAAD, and TGCT. In the correlation analysis of microsatellite instability, FOXP3 is positively correlated with UCEC and THCA, but negatively correlated with DLBC, ESCA, HNSC, KIRP, LIHC, LUSC, OV, SKCM, STAD, TGCT.

FOXP3 immunotherapy response
In Figure 11, in the three immunotherapy cohorts involved in this study, there was no signi cant difference in the expression of FOXP3 between the response group and the non-responder group. The observed trends indicate that it seems that patients with low FOXP3 expression are more effective for immunotherapy.

Discussion
A comprehensive study of the differential expression of FOXP3 in 33 types of cancers, normal tissues and tumor tissues, has found the potential immunotherapy value of FOXP3. FOXP3 not only participates in the regulation of tumor microenvironment, but also participates in the regulation of tumor local immunity. Therefore, we have conducted related research including tumor microenvironment, immune cells, immunomodulators and immunotherapy response. In this study, our goal is to gain a deeper understanding of the potential immunological association between FOXP3 and 33 human cancers. First, we studied the correlation between FOXP3 and clinical parameters and found that only a small number of cancers have signi cant differences in FOXP3 expression with gender, age, and tumor stage, including CESC, CHOL, COAD, HNSC, KIRC, LUAD, PRAD, SKCM, UCEC. Amaral MGD et al. found that the age difference between FOXP3 expression and oral tongue squamous cell carcinoma was not signi cant [15]. Another study showed that in colorectal cancer, the expression of FOXP3 is related to gender and Dukes staging [16], which is consistent with the results of this study. FOXP3 has a certain prognostic value in certain cancers, and related studies have also shown that the high density of FOXP3 in tumor tissues is a powerful independent prognostic marker related to mortality [17]. Many previous studies have considered FOXP3 as an independent factor affecting the poor prognosis of various cancers. For example, in patients with tongue squamous cell carcinoma and oropharyngeal squamous cell carcinoma, high expression of FOXP3 is associated with low overall survival rate [18,19] . In breast cancer, the expression of FOXP3 in breast cancer tissue is signi cantly higher than that in normal breast tissue, and the increase in FOXP3 expression in tumor tissue is related to poor prognosis [20,21]. The above content illustrates the usefulness of FOXP3 in the prognosis of cancer. Therefore, we hypothesize that the regulation of FOXP3 expression in various types of cancer may be clinically bene cial.
In order to further study the potential value of FOXP3, the correlation between FOXP3 and immune cell in ltration was discussed. In terms of immune cell in ltration, FOXP3 expression is positively correlated with plasma cell content in ACC and UVM. FOXP3 expression is positively correlated with the content of T cells follicular helper in ACC; FOXP3 expression is positively correlated with T cells CD8 in KICH, SARC, UVM. Previous studies have shown that CD8 and TGF-β1 are the two main factors in the tumor immune microenvironment. CD8+ T cells are the most important effector executive cells in the tumor immune microenvironment. Inhibiting their activity will affect the body's immune defense function. FOXP3 can help tumor cells escape by inhibiting or killing CD8+ T cells[ twenty two]. In KICH and CHOL, we observed that FOXP3 expression is negatively correlated with NK cells activated. In the analysis of various immunosuppressive agents, we found that the expression of FOXP3 was correlated with the TIGIT of TGCT, and negatively correlated with the KDR of TGCT. TIGIT is an immunosuppressive receptor of T cell immunoglobulin, which is mainly expressed in T cells and natural killer cells, and can inhibit the function of immune cells through a variety of mechanisms to affect the prognosis of cancer patients [23,24]. Based on these data, we propose that there may be potential mechanisms for FOXP3, TIGIT and T cells, but this requires further experimental veri cation. In the analysis of immunostimulants and MHC molecules, the expression of FOXP3 was positively correlated with ICOS and HLA-DRA of TGCT, but negatively correlated with CD276. This discovery may lead to CD276 becoming a new type of immunomodulator. Considering the relationship between FOXP3 and THCA, BRCA, TGCT, ESCA, LGG, we conducted a GSEA analysis to study the relationship between FOXP3 signaling pathways and these cancers. The results showed that the pathway CELL ADHESION MOLECULES CAMS was enriched in the FOXP3 high expression group in ESCA and THCA. There have been few reports on FOXP3 and this pathway. The discovery of this pathway may provide some ideas for immunotherapy research. In addition, in this study, we found that tumor mutation burden and microsatellite instability are shown to be related to FOXP3 in some cancers. Tumor mutation load can be used to predict the e cacy of immune checkpoint blockade and the evaluation of tumor neoantigen load, and it has become a biomarker in certain cancer types [25]. Microsatellite instability is a hallmark feature of DNA mismatch repair defects, and it is an important diagnostic and prognostic cancer biomarker [26]. FOXP3 is positively correlated with TMB and MSI in UCEC, and negatively correlated with two immunotherapy biomarkers in DLBC, HNSC, and LUSC. This indicates that FOXP3 has an indirect effect on the immunotherapy of UCEC, DLBC, HNSC, and LUSC. Finally, we studied the correlation between FOXP3 and the three immunotherapy cohorts. However, no signi cant difference was found in the three immunotherapy cohorts. We hypothesized that FOXP3 may be used to image immunotherapy response through TICIT, ICOS, HLA-DRA, etc. . Our research only involves three immunotherapy cohorts, and we cannot fully clarify the actual immunotherapy value of FOXP3. In the future, more immunotherapy cohorts should be studied.
This article studies the value of FOXP3 in 33 cancer types, provides valuable insights into the role of FOXP3 in cancer immunotherapy, and explores FOXP3 and important immunological indicators (immune cell in ltration, immunomodulators, and immune biology Markers). These ndings may help to understand the potential link mechanism between FOXP3 and immunotherapy. Not all of the 33 human cancers in this study showed an association with FOXP3 in the tumor immune microenvironment, indicating that FOXP3 only plays an immunological role in speci c cancer types. These ndings will help immune targeted therapy Research. However, considering that bioinformatics analysis was used in this study, the ndings of the study are only preliminary. We also need to follow up with more relevant experimental veri cation.

Conclusions
This study is the rst to explore the immunotherapy value of FOXP3 in 33 human cancers, investigate the value of FOXP3 in 33 cancer types, provide valuable insights into the role of FOXP3 in cancer immunotherapy, and explore the role of FOXP3 and FOXP3 in cancer immunotherapy. Important immunological indicators (immune cell in ltration, immunomodulators and immunobiological markers). We believe these ndings will help provide new insights into tumor immunotherapy research.   Figure 1 Shows the analysis and indicators used in this study. In the clinically relevant part, differential FOXP3 expression analysis was performed between different tissues (tumor and normal), age (<=65 and >65), gender (male and female), and stage (I, II, III, IV). Kaplan-Meier was used to perform correlation analysis of survival analysis. In the part of immune mechanism, GSEA explored related signal pathways based on the expression of FOXP3. In the immunotherapy response part, Wilcoxon test was performed according to the expression of FOXP3 in the response group and the non-reaction group.

Figure 2
Clinical correlation analysis of FOXP3. (A) shows the differential expression analysis of FOXP3 tumor group and normal group in 33 cancers; (B) represents the correlation between age and FOXP3; (C) shows the correlation between tumor stage and FOXP3; (D) shows The correlation between gender and FOXP3. "*" means p <0.05, "**" means p <0.01, and "***" means p <0.001.

Figure 3
Kaplan-Meier survival analysis. The relationship between FOXP3 and the overall survival of these cancers.
Page 15/19 Figure 4 Kaplan-Meier survival analysis. The relationship between FOXP3 and these cancers in disease-free survival, disease-speci c survival, and progression-free survival.   Correlation between FOXP3 expression and immune cell in ltration. The CIBERSORT algorithm was used to calculate the immune cell in ltration, and the results of R>0.5 and p<0.05 were retained.

Figure 7
The correlation between FOXP3 and the expression of immunosuppressive agents. Red indicates positive correlation, and blue indicates negative correlation. Take the top 2 strongest associations and display them through dot diagrams.

Figure 8
Page 18 /19 The correlation between FOXP3 and the expression of immunostimulants. Red indicates positive correlation, and blue indicates negative correlation. Take the top 2 strongest associations and display them through dot diagrams.

Figure 9
The correlation between FOXP3 and MHC molecule expression. Red indicates positive correlation, and blue indicates negative correlation. Take the top 2 strongest associations and display them through dot diagrams.

Figure 10
GSEA results based on KEGG database. In each graph, the curve on the left indicates that the pathway is enriched in the high FOXP3 expression group, while the curve on the right indicates that the pathway is enriched in the low FOXP3 expression group.