Identification of the Prognosis Value and Potential Mechanism of Immune Checkpoints in Renal Clear Cell Carcinoma Microenvironment

Background Kidney Renal Clear Cell Carcinoma (KIRC) is one of the most prevalent types of cancer worldwide. KIRC has a poor prognosis and, to date, immunotherapy based on immune checkpoints is the most promising treatment. However, the role of immune checkpoints in KIRC remains ambiguous. Methods Bioinformatics analyses and qRT-PCR were performed to explore and further confirm the prognostic value of immune checkpoint genes and their correlation with immune infiltration in KIRC samples. Results The expression of the immune checkpoint genes CD274, PDCD1LG2, HAVCR2, CTLA4, TIGFT, LAG3, and PDCD1 was upregulated in KIRC tissues. These genes were involved in the activation of the apoptosis pathway in KIRC. Low expression of CD274 and HAVCR2 and high expression of CTLA4 were associated with poor overall survival (OS), progression-free survival (PFS), and disease-free survival (DFS) of KIRC patients. The univariate and multivariate analyses revealed that CTLA4, HAVCR2, age, pTNM stage, and tumor grade were independent factors affecting the prognosis of KIRC patients. A predictive nomogram demonstrated that the calibration plots for the 3‐year and 5‐year OS probabilities showed good agreement compared to the actual OS of KIRC patients. The expression of CTLA4 and HAVCR2 were positively associated with immune cell infiltration, immune biomarkers, chemokines, and chemokine receptors. Moreover, miR-20b-5p was identified as a potential miRNA target of CTLA4 in KIRC. Conclusion Our study clarified the prognostic value of several immune checkpoint regulators in KIRC, revealing a CTLA4/miR-20b-5p axis in the control of immune cell infiltration in the tumor microenvironment.


INTRODUCTION
Renal cell carcinoma (RCC) is one of the most prevalent malignant tumors worldwide, accounting for 2.4% of all cancers (1). In the world, over 403,000 people are initially diagnosed with RCC every year, and 175,000 patients will die of this disease (2). Clear cell renal cell carcinoma (KIRC or ccRCC) is the most frequent histological subtype of RCC, and it accounts for most cancer-related deaths (3,4). Due to the lack of significant clinical symptomatology, KIRC can remain clinically occult, and therefore patients are initially diagnosed in an advanced TNM stage. Late diagnosis generally correlates with lower survival, leading to a poor 5-years survival rate in KIRC patients (5). Moreover, the mortality rates of KIRC have risen to high levels and stabilized in the past ten years (6). The 5-year disease specific survival of RCC patients in stage I was about 80-95%, while it dropped sharply to less than 10% in KIRC patients in stage IV, whose median overall survival was only 10-15 months (4).
Latest studies indicated that immunotherapy based on immune checkpoint regulators is the most promising treatment for KIRC, especially in advanced stages (7). Thus, it is vital to clarify the relation between KIRC and immune infiltration, as well as to identify immune-associated mechanisms and markers for the prognosis and therapy of KIRC. Immune checkpoint molecules secreted from immune cells will inhibit the function of immune cells so that the body cannot produce an effective anti-tumor immune response, leading to immune escape and tumor formation (8). Previous studies have identified many immune checkpoints, including but not limited to SIGLEC15, CD274 (PD-L1), HAVCR2 (TIM-3), PDCD1, CTLA4, LAG3, PDCD1LG2, and TIGIT (8,9). The PD-1 (CD279) inhibitor nivolumab improved overall survival in advanced renal cell carcinoma following prior anti-angiogenic therapy, suggesting immunotherapy as a promising strategy for the management of KIRC (10). Another meta-analysis revealed that PD-1/PD-L1 agents showed a better performance in the treatment for sarcomatoid renal cell carcinoma than sunitinib (11). However, immune checkpoint inhibitors can cause various immune-related adverse events, including adrenal insufficiency and autoimmune hepatitis (12). Thus, further studies should investigate the relationship between these immune checkpoints and their role in the prognosis and management of KIRC.
With the continuous development of gene sequencing technologies and the establishment and improvement of various tumor databases, bioinformatic research has been suggested as one of the most reliable ways to accelerate clinical and translational cancer research. Our study aimed to clarify the correlation between immune checkpoint expression, immune infiltration, and KIRC prognosis. Moreover, our results might provide additional data about the molecular mechanism of immune checkpoint regulators in immune infiltration.

Datasets
In order to explore the clinical significance of immune checkpoints in KIRC, we first retrieved the KIRC gene expression profile from the Cancer Genome Atlas (TCGA) database (https://portal.gdc. cancer.gov/) and Oncomine (https://www.oncomine.org/). In TCGA, the KIRC dataset (TCGA-KIRC) was downloaded for analysis, and the corresponding clinical information, including gender, tumor grade, and survival status of the patients, were also downloaded and sorted. The Oncomine datasets GSE14994, GSE6344, and GSE11151, were also downloaded to analyze the expression of immune checkpoints.

Gene Expression Analysis
The gene expression of eight immune checkpoints was analyzed using Oncomine and TCGA. In TCGA, the relevant transcripts and expression values of the genes were extracted and visualized using the R software packages "ggplot2" and "pheatmap" (R foundation for statistical computing, 2020; version 4.0.3). In Oncomine, Student's t-test was used to evaluate the significance and compare the gene expression of immune checkpoints between normal tissues and KIRC tissues. A p-value threshold of 0.05 and a fold-change of 2 were set to define statistically significant changes.

Genetic Mutation, Drug Sensitivity, and Cancer-Related Pathway Analysis
The genetic mutation data were downloaded from the TCGA dataset, and the genetic mutation of eight immune checkpoints was analyzed and visualized using the "maftools" R package. We collected 265 small molecules from Genomics of Drug Sensitivity in Cancer (GDSC) to analyze the correlation between immune checkpoints and drug sensitivity. Pearson correlation analysis was used to calculate the correlation coefficient, and a p-value < 0.05 was considered statistically significant. The method of cancer-related pathway analysis has been applied as in Ye et al. (13). Immune checkpoint expression was divided into two groups, High and Low, with median expression. The significant difference of pathway activity score (PAS) between groups is evaluated by the Student's t-test. We considered that an immune checkpoint might have an activating effect on a pathway when PAS H (immune checkpoint group High) > PAS L (immune checkpoint group Low); otherwise, the checkpoint might have an inhibitory effect on the pathway.

Enrichment Analysis and Protein-Protein Interaction (PPI) Networks of Immune Checkpoints
The functional annotation of immune checkpoints was analyzed with Metascape (https://metascape.org/), a powerful tool to comprehensively analyze and interpret OMICs-based data (14). We also constructed a PPI network of immune checkpoints using GeneMANIA (https://genemania.org/), a prediction server for gene prioritization and predicting gene function (15).

Prognosis Analysis
The Kaplan-Meier method was applied to analyze the prognosis of immune checkpoints in KIRC. p-value and hazard ratio (HR) with 95% confidence interval (CI) were calculated using a log-rank test. A predictive nomogram was constructed based on proper terms identified by univariate and multivariate cox regression analysis. A forest was used to show the p-value and HR as well as 95% CI of each variable using the "forestplot" R package.

Clinical Tissues and qRT-PCR
We obtained 30 KIRC tissues and normal kidney tissues from patients who underwent tumor resection in the Affiliated Hangzhou First People's Hospital. Histological diagnosis and tumor grade were assessed by three experienced pathologists following the 2010 American Joint Committee on Cancer (AJCC) staging system. All procedures were approved by the Ethics Committee of Affiliated Hangzhou First People's Hospital, and informed consent was obtained from each patient.
Total RNA of clinical tissues was extracted using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc), and PrimeScript RTpolymerase (Vazyme) was used to synthesize the cDNA according to the manufacturer´s instructions. RT-qPCR was performed with SYBR-Green Premix (Qiagen GmbH) with specific PCR primers (Sangon). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. The primers of GAPDH and immune checkpoints were shown in Supplementary Table 1. The fold-change was calculated as previously described with the 2 −DDCt method. The Student's t-test was conducted to compare the expression of immune checkpoints in KIRC and normal tissues. Kaplan-Meier analysis was performed to evaluate the prognosis value of immune checkpoints in KIRC.

Immune Infiltration Analysis
Immune infiltration of immune checkpoints in KIRC was analyzed using TIMER (https://cistrome.shinyapps.io/timer/), a web server designed for comprehensive analysis of tumorinfiltrating immune cells (16). Spearman's correlation analysis was performed to explore the relationship between immune checkpoints and immune cell infiltration and the expression of immune biomarkers, chemokines, and chemokine receptors. A p-value of less than 0.05 was considered statistically significant.
Construction of miRNA-mRNA Regulatory Axis miRNAs binding to immune checkpoint genes were identified using starBase v3.0. The expression and prognosis value of miRNA in KIRC were explored as described above.

Cancer Hallmarks Analysis of Immune Checkpoints in KIRC
In order to investigate the role of immune checkpoint molecules in KIRC, we also performed cancer hallmarks analysis. The  Figure 2A). HAVCR2 was the most commonly altered gene among all the studied immune checkpoint genes, and about 19% of the total TCGA-KIRC cases counted with a HAVCR2 genetic mutation (Figure 2A).
Activation and inhibition of cancer hallmark pathways play a vital role in tumorigenesis and progression. Therefore, we then explored the effect of these immune checkpoint regulators in several cancer hallmark pathways in KIRC. These pathways included TSC/mTOR, RTK, RAS/MAPK, PI3K/AKT, Hormone ER, Hormone AR, EMT, DNA Damage Response, Cell Cycle, and Apoptosis pathways. The results indicated that  immune checkpoints were involved in the activation of the apoptosis pathway, EMT pathway, and the inhibition of DNA damage response pathway in KIRC ( Figure 2B). In order to identify potential therapeutic targets, a critical step is to evaluate the relation between immune checkpoints and existing drug targets. Interestingly, a drug sensitivity analysis revealed that most of these immune checkpoints are sensitive to most of the small molecules or drugs from GDSC (negative correlation, coefficient from -0.50 to -0.10) ( Figure 2C, p<0.05). Moreover, co-expression analysis suggested a moderate to high correlation (coefficient from 0.25 to 0.80) among several immune checkpoint molecules ( Figure 2D, p<0.05).

Enrichment Analysis of Immune Checkpoints in KIRC
In order to clarify the immune checkpoint-associated functions in KIRC, we performed a gene enrichment analysis. As shown in Supplementary Figure 2B, these immune checkpoints were mainly associated with biological adhesion, immune system process, regulation of the biological process, cell proliferation, and cellular process in Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses (p<0.05). Moreover, the PPI network based on immune checkpoints suggested that these immune checkpoints were mainly associated with T cell costimulation, lymphocyte costimulation, regulation of T cell, and lymphocyte activation (Supplementary Figure 2B, p<0.05).

Prognosis Value of Immune Checkpoints in KIRC
Next, we explored the prognosis value of immune checkpoints in KIRC. The results suggested that CTLA4, HAVCR2 and CD274 were significantly associated with the overall survival (OS), progression free survival (PFS), and disease-free survival (DFS) of KIRC patients (  Figure 4). These data demonstrated that CD274, HAVCR2 and CTLA4 might serve as prognostic biomarkers in KIRC.
Considering clinicopathologic features and HAVCR2 and CTLA4 as potential prognostic biomarkers, we constructed a predictive nomogram to predict the 1-year, 3-year, and 5-year OS rates in the discovery group using the cox regression algorithm. The nomogram demonstrated that the predicted calibration plots for 3-and 5-year OS probabilities showed good agreement compared to the actual OS of KIRC patients ( Figures 4C, D, p<0.001).

The Correlation Between CTLA4/HAVCR2 and Clinical Characteristics in KIRC
The above results revealed that CTLA4 and HAVCR2 were independent factors affecting the prognosis of KIRC and were associated with the OS, PFS, and DFS of KIRC patients. Therefore, we selected CTLA4 and HAVCR2 for further analysis. In order to explore the function of CTLA4 and HAVCR2 in KIRC, we analyzed the correlation between CTLA4 and HAVCR2 expression and the clinical characteristics of KIRC. As a result, KIRC patients with a high pT stage (p = 0.00023) had a low CTLA4 expression compared with those with a low pT stage ( Figure 5A). Moreover, KIRC patients with a high pN stage (p = 0.04) had a lower HAVCR2 expression than those with a low pN stage ( Figure 5B).

Validation of the Expression and Overall Survival of CTLA4 and HAVCR2 in KIRC
We then performed qRT-PCR to further confirm the expression of CTLA4 and HAVCR2 in KIRC. As expected, the relative mRNA level of CTLA4 (p < 0.001, Figure 6A) and HAVCR2 (p < 0.001, Figure 6B) in KIRC samples was elevated compared to normal renal tissues. In addition, we also studied the overall survival of these patients. The data indicated a poor overall survival in those patients with high CTLA4 expression (p = 0.040, Figure 6C) and low HAVCR2 expression (p = 0.027, Figure 6D), consistently with previous results.

The Correlation Between CTLA4/HAVCR2 and Immune Infiltration in KIRC
The immune infiltration grade is an independent predictor of sentinel lymph node status and survival in cancers (4,18,19).  Table 2). A previous study revealed that chemokines and their receptors play a vital role in immune infiltration (20). In our study, we found that the expression of CTLA4 and HAVCR2 was positively correlated with the expression chemokines as well as chemokines receptors ( Figures 7C, D). This evidence indicated a possible association between CTLA4/HAVCR2 and immune infiltration in KIRC patients.

DISCUSSION
Previous studies revealed that immune checkpoint regulators were correlated with many biological processes, thus affecting the clinical outcomes of cancer patients (21,22). It is suggested that immune checkpoints could act as markers to predict the prognosis of many cancers, including renal cell carcinoma, lung cancer, and breast cancer (23)(24)(25). However, the prognostic value of immune checkpoints and their association with immune infiltration of KIRC remain unclear. Thus, this study aimed to detect the role of these immune checkpoints in the prognosis and immune infiltration of KIRC. We first explored the expression level of several immune checkpoint molecules in KIRC samples. We found that the mRNA level of most of these immune checkpoints, including CD274, PDCD1LG2, HAVCR2, CTLA4, TIGFT, LAG3, and PDCD1, was altered in KIRC patients. These immune checkpoints might play a vital role in the oncogenesis and progression of KIRC. As expected, further analysis revealed that these immune checkpoints were involved in the activation of the apoptosis pathway in KIRC. Therefore, immune checkpoints may inhibit oncogenesis and progression by activating the apoptosis pathway. Moreover, our study also found that CTLA4 and HAVCR2 acted as prognostic biomarkers in KIRC and were associated with overall survival (OS), progression-free survival (PFS), and disease-free survival (DFS) of KIRC patients. In agreement with our results, previous studies have suggested certain immune checkpoints as a prognostic biomarker of many cancers. HAVCR2 was a prognostic biomarker for gastric cancer and was negatively associated with OS (26), and two independent studies revealed that HAVCR2 was a diagnostic and prognostic biomarker of osteosarcoma (27) and large B-cell lymphoma (28).
Another important finding of our study is that a significant correlation was obtained between the expression of CTLA4 and HAVCR2 and immune cells, immune biomarkers, chemokines, and chemokine receptors. All these factors play a critical function in controlling tumor immune infiltration, anticancer immunity, and other biological processes, thus affecting the prognosis of the patients. For example, previous studies indicated that tumorinfiltrating CD8+ T cells determined poor prognosis and immune evasion (29) and B cells predicted dismal survival and worse treatment response in KIRC (30). Another bioinformatics study suggested that low mRNA levels of the chemokines CXCL1/2/3/5/13 were associated with a significantly better prognosis in KIRC (4).

CONCLUSION
Our study performed a comprehensive analysis of the prognostic value of immune checkpoints in KIRC and their association with

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Affiliated Hangzhou First People's Hospital. The patients/participants provided their written informed consent to participate in this study.

AUTHOR CONTRIBUTIONS
GL performed data analysis work and aided in writing the manuscript. YW designed the study, assisted in writing the manuscript. PW edited the manuscript. All authors contributed to the article and approved the submitted version.