T-Cell Exhaustion-Related Gene Signatures Predict Clinical Outcomes and Immunotherapy Response in Kidney Renal Clear Cell Carcinoma

doi:10.21203/rs.3.rs-4802145/v1

Download PDF

Article

T-Cell Exhaustion-Related Gene Signatures Predict Clinical Outcomes and Immunotherapy Response in Kidney Renal Clear Cell Carcinoma

https://doi.org/10.21203/rs.3.rs-4802145/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

T-cell exhaustion (TEX) is a crucial immune escape mechanism and a pathway of resistance in cellular immunotherapy, yet its prognostic role in kidney renal clear cell carcinoma (KIRC) remains unclear. This study analyzed 518 KIRC patients from the TCGA dataset, identifying TEX-related genes through Gene Set Variation Analysis (GSVA) and Weighted Gene Co-Expression Network Analysis (WGCNA). Using survival random forest and LASSO-Cox analyses, eight differentially expressed genes (RUFY4, NOD2, IL15RA, CXCL13, GBP5, DERL3, SPIB, and SLCO5A1) were selected to construct a TEX risk model. Functional analyses, including GO, KEGG, GSEA, CIBERSORT, and ssGSEA, explored the relationship between TEX risk scores and signaling pathways and immune cell infiltration. The IMvigor210 dataset evaluated the correlation between TEX risk scores and immunotherapy response, while single-cell data analysis and qRT-PCR validated the expression of a key TEX gene. The TEX risk model demonstrated accurate prognostic prediction for KIRC patients, serving as a new independent prognostic factor. GSEA results highlighted the enrichment of tumor proliferation, migration, and immunity functions within the model. TEX features were associated with immune cell infiltration and specific immune checkpoints, effectively predicting clinical responses to immunotherapy. Thus, TEX signatures are pivotal in clinical decision-making for KIRC, helping to distinguish patients and guide treatment strategies for maximum benefit.

Biological sciences/Cancer/Urological cancer

Biological sciences/Cancer/Cancer therapy/Cancer immunotherapy

Kidney Renal Clear Cell Carcinoma

T-cell Exhaustion

Immunotherapy

prognosis

tumor microenvironment

immune checkpoint inhibitors

According to the Global Cancer Statistics 2020, kidney cancer is the 14th most common malignancy worldwide, with an estimated 431,288 new cases in 2020¹. Due to significant gender differences in incidence, kidney cancer ranks as the 9th most common cancer among men and the 14th among women¹. In 2020, 179,368 people globally died from kidney cancer (115,600 men and 63,768 women), with a global age-standardized rate (ASR) of 1.8 per 100,000¹. Due to its insidious nature, kidney cancer is not sensitive to radiotherapy and chemotherapy. Even with surgery or ablation therapy, more than one-third of cases experience recurrence or metastasis². Over the past decade, RCC treatment has undergone revolutionary changes, from high-dose cytokine therapy combined with surgical tumor resection to extensive stage-dependent treatment regimens based on targeted therapies, particularly anti-angiogenic drugs targeting the vascular endothelial growth factor (VEGF) pathway and cellular immunotherapies. Among these, immune checkpoint inhibitors (ICIs) and adoptive cell transfer are relatively effective cellular immunotherapies³.

T-cell exhaustion (TEX) is a state of T-cell dysfunction that occurs during chronic inflammatory infections and malignancies, characterized by the high expression of inhibitory cytokines, reduced cytotoxicity, and proliferation capacity of CD8 + T cells, ultimately leading to a lack of cascade responses⁴. It is generally believed that the efficacy of cancer immunotherapy is related to the strength of T-cell function, making T-cell exhaustion a pathway of resistance in cellular immunotherapy⁵.

Kidney renal clear cell carcinoma (KIRC) is a highly immunogenic tumor. However, the high infiltration of immunosuppressive cells (such as regulatory T cells and myeloid-derived suppressor cells) in its tumor microenvironment (TME) leads to immune dysfunction⁶. Single-cell transcriptomics analysis has shown that CD8 + T cells in the TME exhibit a highly exhausted state⁷. The key mechanism of immune checkpoint inhibitors is to reduce T-cell exhaustion in the TME. Evidence indicates that inhibiting the PD-1 inhibitory receptor pathway can suppress TEX and activate immune anti-tumor effects⁴. Therefore, immune checkpoint blockade and its combination therapies have become standard treatments for patients with advanced RCC.

To identify KIRC patients with different TEX characteristics and find TEX-related genes with high prognostic value, this study used various analytical techniques to construct a prognostic signature of TEX-related genes, elucidate the relationship between TEX process-related genes and KIRC prognosis, and explore the association of this prognostic signature with immune infiltration and immunotherapy outcomes in KIRC patients.

2.1 Data Collection

2.2 TEX Pathway Activity Enrichment Score

We referred to studies on TEX, TEX signaling pathway signal transduction, and marker genomes in the Molecular Signatures Database⁹(MSigDB, V7.2) (https://www.gsea-msigdb.org/gsea/msigdb) and used four pathways—T cell receptor, IL2, IFN-γ, and TNFα to represent the TEX signaling pathway. TEX pathway activity scores for each patient were estimated using the "GSVA"¹⁰ R package. Additionally, TEX-related data on KIRC immune cell infiltration were collected from the ImmuCellAI online database¹¹ (http://bioinfo.life.hust.edu.cn/).

2.3 Weighted Gene Co-Expression Network Analysis (WGCNA)

We utilized the systems biology method of Weighted Correlation Network Analysis¹² (WGCNA) to describe the correlation patterns among genes. Differentially expressed genes in KIRC were analyzed using the WGCNA R package. To ensure the topology network was scale-free, the soft threshold parameters were set at a power of 7 and a scale-free R² of 0.9. Seven modules were identified, with the green module showing the closest relationship to TEX, warranting further investigation.

2.4 Construction and Validation of the TEX Signature

To identify TEX-related genes of significant prognostic value, we first performed univariate Cox regression analysis on the green module genes. Genes with P < 0.01 in the univariate Cox analysis were further screened using the Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis. Additionally, we performed random forest analysis on genes selected from the univariate Cox analysis, with the initial nTree set to 10,001. The top 20 genes in relative importance were analyzed, and eight genes were identified by both algorithms. Risk scores for each patient were then calculated using the formula: Risk score = ExpressionmRNA1 × CoefmRNA1 + ExpressionmRNA2 × CoefmRNA2 + ... + ExpressionmRNAn × CoefmRNAn. Patients in the training set were divided into high-risk and low-risk groups based on the optimal cutoff value calculated using surv_cutpoint. Kaplan-Meier survival analysis and Principal Component Analysis (PCA) were conducted on the low TEX_score and high TEX_score groups. One-year, three-yea

, and five-year Receiver Operating Characteristic (ROC) curve analyses were performed using the "survival," "survminer," and "timeROC" R packages. The same Kaplan-Meier survival and ROC curve analyses were performed on the training cohort.

2.5 Gene Set Enrichment Analysis (GSEA)

GSEA was conducted using the Hallmark, Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and GSEA reference gene sets to analyze biological pathway enrichment between the two groups.

2.6 Establishment and Validation of the Nomogram Scoring System

Univariate analysis was first performed on the clinical pathological characteristics and ICD_score of the TCGA cohort. Multivariate analysis was also conducted, and a predictive nomogram was developed using the "rms" R package. Decision Curve Analysis (DCA) curves were used to describe the consistency between predicted survival events and actual observations.

2.7 Tumor Immune Prediction Significance of the TEX Risk Model

Heatmaps and box plots illustrated the expression of tumor immunosuppressive genes and immune checkpoints between different risk groups. The CIBERSORTX algorithm analyzed the infiltration of 22 different types of immune cells. Using the IMvigor210 package, we determined the gene expression and immunotherapy efficacy in the IMvigor210 cohort, widely used to analyze the effectiveness of immunotherapy⁸.

2.8 Cell lines

All cells were cultured in a 37°C incubator in an atmosphere of 5% CO2. The normal human kidney cell line HK-2 and the renal cancer cell line 786-O were from the Chinese Academy of Sciences (Shanghai, China). Cell culture medium, plates, and dishes were from SaveWill (Wuhan, China). HK-2 and 786-O were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin.

2.9 RNA extraction and qRT-PCR:

Total RNA was extracted using the EasyPure® Fast Cell RNA Kit (TransGen Biotech, Beijing). The concentration of RNA was measured using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific Inc., USA). Total RNA (2 µg) was reverse transcribed into cDNA using the FastKing RT Kit (TIANGEN Biotech. Co., Beijing). According to the manufacturer's instructions, real-time quantitative PCR was performed using the SuperReal PreMix Plus (TIANGEN Biotech. Co., Beijing) and the StepOnePlus Real-Time PCR System. The relative expression levels of target genes were calculated using the 2^-△△CT method. The sequences of the primers were as follows:

DERL3_F:CTACTTCCTGGAGGACGTCTTC

DERL3_R:TCTGCTGTGTCCAGATGGTCAG

2.10 Statistical Analysis:

Statistical analyses were performed using R software (version 4.3.3). Differentially expressed genes in KIRC tissues and normal tissues were analyzed using the "edgeR" package. Cox regression analysis was used to screen genes related to prognosis. The Kaplan-Meier method tested the overall survival (OS) of patients. ROC curve analysis was performed using the "time-ROC" package. The "rms" package was used to plot the nomogram. Correlations between risk scores, prognostic genes, and clinicopathological variables were compared using the Student's t-test or non-parametric Mann-Whitney rank-sum test. P < 0.05 was considered statistically significant.

3.1 Identification of T Cell Exhaustion-Related Genes

We first performed differential expression analysis on the TCGA-KIRC cohort using the edgeR package, identifying a total of 4759 differentially expressed genes (DEGs). Analysis of GSE139555 in the TISCH database¹³ revealed that TEX is associated with T CELL RECEPTOR (Fig. 1A-C). Additionally, research has shown that TEX is related to IL2, IFN-γ, and TNFα pathways ¹⁴. Therefore, we obtained enrichment scores for four TEX-related pathways for each sample using GSVA and directly obtained T cell exhaustion enrichment scores from the ImmuCellAI online database. For these DEGs, we used WGCNA to identify key modules most strongly associated with T cell exhaustion progression in the TCGA-KIRC cohort. During the construction of the co-expression network, we observed a soft threshold power β of 7 when the scale-free topology fitting index reached 0.9 (Fig. 1D, E). By merging similar modules, we ultimately obtained 14 color modules and found that the MEgreen group was most significantly associated with T cell exhaustion scores (p < 0.01). Therefore, we selected the MEgreen module as the key module and obtained 253 candidate TEX-related genes (Fig. 1F-H).

3.2 Construction of the TEXScore Predictive Model

We first screened candidate TEX-related genes using univariate COX analysis, obtaining 53 prognosis-related genes (P < 0.01). To exclude co-expressed TEX genes and avoid overfitting, we further screened these genes using LASSO regression analysis, ultimately identifying 16 genes (Fig. 2A, B). Additionally, survival random forest analysis identified 20 genes most related to clinical outcomes (Fig. 2C). Eight genes were identified by both algorithms: RUFY4, NOD2, IL15RA, CXCL13, GBP5, DERL3, SPIB, and SLCO5A1 (Fig. 2D). The model was constructed using the LASSO regression algorithm, calculated as risk score = (0.75578 x RUFY4 exp) + (0.80200 x NOD2 exp) + (0.46737 x IL15RA exp) + (0.24353 x CXCL13 exp) + (-0.83225 x GBP5 exp) + (0.04164 x DERL3 exp) + (0.34617 x SPIB exp) + (1.97911 x SLCO5A1 exp). RUFY4, NOD2, IL15RA, CXCL13, DERL3, SPIB, and SLCO5A1 were significantly positively correlated with the risk score, with SLCO5A1 showing the strongest positive correlation. Conversely, GBP5 was significantly negatively correlated with the risk score. To demonstrate the model's stability and reliable generalization, we used the caret R package to split the TCGA-KIRC data into training and test cohort. The risk scores for each sample in the training and validation cohorts were calculated using the same risk formula. Based on the optimal risk score cutoff value calculated using surv_cutpoint, patients were divided into high-risk (HR) and low-risk (LR) subgroups to explore the prognosis differences between HR and LR groups (Fig. 2E). We found that as the risk increased in KIRC patients in both cohorts, patients exhibited a survival disadvantage with reduced OS and increased mortality (Fig. 2F, G). Kaplan-Meier curves showed significant prognosis differences between HR and LR patients in both cohorts, with LR patients having a more pronounced survival advantage (Fig. 2H, I). The time-ROC curve was used as a tool to predict 1-year, 3-year, and 5-year survival times, with the training cohort's AUCs being 0.75, 0.70, and 0.74, respectively. The test cohort's AUCs were 0.60, 0.64, and 0.66, respectively (Fig. 2J, K). This indicates the model has excellent predictive performance.

3.3 Creation of a TEX Feature-Based Nomogram Incorporating Clinical Characteristics

To validate the reliability and clinical value of the TEX-based biological feature as a prognostic predictor, we compared the risk score of each KIRC patient with clinical indicators (age, gender, stage) and observed the correlation between each factor and patient prognosis through univariate and multivariate Cox analyses. The results showed that stage and risk score (P < 0.001) were significant prognostic factors in univariate Cox analysis (Fig. 3A). Multivariate Cox analysis also indicated that stage and risk score (P < 0.001) were significant prognostic factors (Fig. 3B). Therefore, the TEX feature-based risk score can serve as an independent prognostic factor related to KIRC patient prognosis. To quantitatively predict patient prognosis and provide information for clinical decision-making, we integrated the risk score and clinical indicators to construct a nomogram as a tool to predict 1-year, 3-year, and 5-year survival probabilities (Fig. 3C). Calibration analysis showed that the OS prediction curves for 1, 3, and 5 years were highly similar to the ideal 45-degree calibration line, indicating excellent stability of the nomogram (Fig. 3D). Subsequently, DCA showed that the nomogram and risk score produced greater net benefits and predictive benefits, indicating that the model's risk score and nomogram could serve as major decision-making factors (Fig. 3E).

3.4 Functional Enrichment Analysis Between HR and LR Groups

We performed functional enrichment analysis between HR and LR groups in the training set data. GO analysis results included Cellular Component (CC), Molecular Function (MF), and Biological Process (BP) (Fig. 4A). We further explored BP and found that pathways related to immunity, such as regulation of T cell activation, humoral immune response, fatty acid transport, positive regulation of T cell migration, and lymphocyte chemotaxis, were significantly enriched (Fig. 4B). KEGG analysis results showed that pathways such as PPAR signaling pathway, Glutathione metabolism, ECM-receptor interaction, and TCA cycle were downregulated in the HR group (Fig. 4C), while pathways such as TNF signaling pathway, TGF-beta signaling pathway, PI3K-Akt signaling pathway, NF-kappa B signaling pathway, JAK-STAT signaling pathway, IL-17 signaling pathway, HIF-1 signaling pathway, GABAergic synapse, Cytokine-cytokine receptor interaction, and Chemokine signaling pathway were upregulated in the HR group (Fig. 4D). These pathways are inolved in tumor growth, migration, and tumor immunity, which may reveal the potential pathways through which TEX exerts its effects. GSEA analysis results also validated the KEGG results (Fig. 4E).

3.5 TEX Risk Score Predicts Immune Cell Infiltration

To further investigate the role of TEX features in the KIRC tumor immune microenvironment, we quantified immune cell infiltration between LR and HR groups in the training set using the CIBERSORT script, generating a heatmap of patients ranked from lowest to highest risk scores, showing the infiltration level corresponding to each immune cell (Fig. 5A). We found that regulatory T cells, CD8 + T cells, and follicular helper T cells were highly infiltrated in the HR group, while resting mast cells, M2 macrophages, naive B cells, resting NK cells, and resting dendritic cells were lowly infiltrated in the HR group (Fig. 5B). We further validated these findings using test set data and found that M0 macrophages and plasma cells were highly infiltrated in the HR group of the test set (Fig. 5C). Using the ssgsea method, we quantified immune cell infiltration in the training set patients and visualized the relationship between risk scores and tumor immune cells (Fig. 5D). We also compared TEX levels between LR and HR groups in the training set using TEX data calculated from the ImmuCellAI in the TCGA KIRC patients, finding higher TEX levels in the HR group (Fig. 5E). This result may explain why, despite high CD8 + T cell infiltration in the HR group, HR patients had worse prognoses. Finally, we analyzed tumor immune subtypes in the training set patients using TCGA tumor immune subtype data from the ImmuneSubtypeClassifier package¹⁵, finding that the main subtypes in both high-risk and low-risk groups were C3, but the HR group had significantly more C1, C2, and C6 subtypes than the LR group(Fig. 5F). C1 (Wound Healing) had elevated expression of angiogenic genes and a high proliferation rate. C2 (IFN-γ Dominant) had the highest M1/M2 macrophage polarization, a strong CD8 signal, and, together with C6, the greatest TCR diversity. C3 (Inflammatory) is characterized by increased expression of Th17 and Th1 genes, with low to moderate tumor cell proliferation. C6 (TGF-β Dominant) shows the highest TGF‐β signature and high lymphocytic infiltration with well-distributed Type I and Type II T cells. The C5 and C6 immune subtypes in tumors have worse prognoses than the C3 immune subtypes, consistent with the prognosis of high- and low-risk KIRC patients¹⁶. In Thorsson, V. et al.'s study, C1, C2, and C6 were associated with poorer OS in KIRC patients, while C3 was associated with better OS. These results may explain the poorer prognosis of HR patients ¹⁶.

3.6 Prediction and Validation of Immunotherapy Efficacy

Given the immense potential of TEX in immunotherapy, we mapped the expression levels of common immune checkpoints between LR and HR populations¹⁷, finding differential expression of several immune checkpoints between the two groups (Fig. 6A, B). Specifically, targets commonly used in KIRC immunotherapy, such as PDCD1, LAG3, CTLA4, TIGIT, CD28, and CD276, were highly expressed in the HR group (Fig. 6C, D). To test the potential of TEX features in predicting the efficacy of immunotherapy in a real immunotherapy cohort, we analyzed immunotherapy data collected in the IMvigor210CoreBiologies package¹⁸, finding that patients with lower risks had better prognoses (Fig. 6E). Patients responding to immunotherapy had lower risk scores (Fig. 6F, G), and the overall response rate in the LR group was higher than in the HR group (Fig. 6H). These results indicate that the TEX risk score is effective in predicting the efficacy of immunotherapy and the prognosis of patients undergoing immunotherapy.

3.7 Single-Cell Sequencing Analysis and Real-Time Quantitative PCR

We separately analyzed the relationship between key genes in the risk model, tumor prognosis, and TEX. We found that the key gene DERL3 is related to the prognosis of KIRC patients (Fig. 7A) and the TEX score of KIRC patients (Fig. 7B). To further explore potential therapeutic target cell interactions for TEX, 10 cell subtypes (ccrcc, Marco, CD8 + T cell, Renal tube, yT2c, endo, fibro, B cell, Tfh, epi) were identified in the single-cell sequencing dataset (Fig. 7C). We found that DERL3 is most highly expressed in B cells (Fig. 7D, E), which was also validated by the GSE111360 dataset in the TISCH database (Fig. 7F-H). This may suggest that DERL3 influences TEX through B cells and could be a potential therapeutic target. Additionally, we detected DERL3 mRNA levels in normal and tumor cell lines. The results showed that the mRNA expression level of DERL3 in 786-O is much higher than in HK-2 (Fig. 7I). Therefore, we believe that the abnormal expression of DERL3 may promote the proliferation and deterioration of KIRC.

With the advent of immunotherapy, the treatment landscape of kidney renal clear cell carcinoma (KIRC) is undergoing a transformative shift. Various immunotherapies have shown significant potential in KIRC treatment. Combination therapies, including immune checkpoint inhibitors (ICIs) with other modalities such as bispecific antibodies, chimeric antigen receptor (CAR) T-cell therapy, and cancer vaccines, are emerging continuously. Due to the differences in epigenetic states, the complexity and heterogeneity of the tumor microenvironment (TME), current research focuses on rationally designing target combinations, developing predictive biomarkers, and improving T-cell efficacy. T-cell exhaustion (TEX) is defined as a dysfunctional state resulting from prolonged exposure of T-cells to antigens and/or inflammatory signals in chronic infections or cancer ¹⁹. In this state, dysfunctional T-cells (including effector T-cells and memory T-cells) lose their ability to eradicate infections and cancer. By investigating CD8 + T-cell dysfunction and exhaustion in the TME, it might be possible to overcome the TEX barrier and improve the clinical efficacy of immune checkpoint blockade therapy. However, despite the significance of T-cell exhaustion in the progression of various cancers, there is a lack of systematic studies on TEX in KIRC. Therefore, we developed a multi-biomarker model based on TEX-related genes to help clinicians evaluate the prognosis and TME of KIRC patients, providing a theoretical basis for personalized and precision therapy.

We used TCGA-KIRC data, single-cell data analysis, and the GSVA algorithm to obtain the expression profiles of the four pathways most closely related to TEX (T cell receptor, IFNG, TNF, and IL-2). Subsequently, we analyzed differentially expressed genes using WGCNA and performed univariate Cox analysis to identify 53 key genes. Using the LASSO regression algorithm and survival random forest analysis, we ultimately selected eight genes (RUFY4, NOD2, IL15RA, CXCL13, GBP5, DERL3, SPIB, SLCO5A1) to construct the TEX risk model. Among these, RUFY4, a member of the RUFY family, influences autophagy and dynamics in immune cell subsets, playing a crucial role in determining the nature of the cellular response to the immediate immune environment. In KIRC, it serves as a prognostic biomarker, predicting the effectiveness of immunotherapy. NOD2, a nucleotide-binding oligomerization domain-like receptor (NLR), is part of the evolutionarily conserved pattern recognition receptors (PRRs) family, closely related to innate and adaptive immunity. In human cervical squamous cell carcinoma (CSCC), NOD2 enhances the tumorigenic and metastatic potential of CSCC cells through multiple pathways. However, in esophageal cancer, NOD2 inhibits the proliferation of esophageal adenocarcinoma cells through autophagy. IL15RA is typically found in a heterotrimeric receptor with IL2RB and IL2RG. It can bind IL15 with high affinity even in the absence of the β and γ subunits, which is related to regulating the body's adaptive immune response and the activation and maintenance of various lymphocyte populations ^20–22. Studies have reported that the co-expression of IL15RA and IL15 in breast cancer cell lines can promote tumor proliferation, prevent tumor cell apoptosis, and enhance cell migration ²³. CXCL13, initially known as BLC (B lymphocyte chemoattractant) or BCA-1 (B cell-attracting chemokine 1), is a homeostatic chemokine that exerts its functions through its receptor CXCR5, initially isolated from Burkitt Lymphoma and therefore designated Burkitt's lymphoma receptor 1 (BLR1). In renal cell carcinoma, the CXCL13/CXCR5 signaling axis is associated with tumor proliferation, migration, and malignant stages, correlating with poor prognosis. However, the role of the CXCL13/CXCR5 axis in tumor immunity is controversial. Some studies suggest that CXCL13 plays a critical role in the formation of tertiary lymphoid structures (TLS) in different tumor tissues, which promote humoral and cell-mediated immune responses against tumors, contributing to effective antigen presentation, cell activation, and differentiation. On the other hand, the CXCL13/CXCR5 axis might indirectly help malignant cells evade T-cell immunity by inducing the IL-10 pathway or recruiting immunosuppressive MDSC and Treg cells in the TME.

GBP5, a member of the guanylate-binding protein family, is an interferon-inducible GTPase involved in innate immune responses against various pathogens ^24,25. GBP5 promotes radio-resistance, metastatic progression, and PD-L1 expression in oral squamous cell carcinoma ²⁶ and also supports the growth and invasion of mouse glioblastoma ²⁷. Conversely, GBP5 is associated with better prognosis in breast cancer and skin melanoma, with IFNγ-induced chemokine upregulation observed in GBP5-high breast cancer ^28,29. In TNBC cells, GBP5 knockdown inhibits PD-L1 levels and NF-κB activity³⁰. DERL3 (derlin-3) is a member of the Derlin family involved in the ER-associated degradation (ERAD) process in cancer cells. DERL3 acts as a tumor suppressor gene in colorectal and gastric cancers ^31,32 with its loss of function leading to uncontrolled accumulation of aberrant proteins, persistent ER stress, and subsequent tumor progression and metastasis. However, DERL3 exhibits tumor-promoting effects in breast and lung adenocarcinomas^33–35. SPIB, an ETS transcription factor involved in B-cell receptor-mediated signal transduction, plays a significant role in tumor development ³⁶. In colorectal cancer cells, SPIB inhibits tumor progression by activating NF-κB and JNK signaling via MAP4K1³⁷ and promotes TAM recruitment in lung cancer by upregulating CCL4 expression ³⁸. SPIB knockdown in ovarian cancer cells reduces PD-L1 (CD274) expression and inactivates the JAK/STAT pathway, inhibiting immune escape³⁹. SLCO5A1 is an organic anion transporting polypeptide (OATP) in the solute carrier organic anion (SLCO) transporter family^40–42. Elevated OATP5A1 expression has been detected in liver cancer, small cell lung cancer, and renal cancer ^43,44, potentially contributing to chemoresistance in small cell lung cancer to Satraplatin ⁴⁴.

Using the constructed TEX signature, we divided the training and test set data into high-risk (HR) and low-risk (LR) groups. Significant prognostic differences between the two groups in the training set demonstrated the independent predictive value of the TEX signature in KIRC. ROC and calibration curve analyses confirmed the superior predictive performance of the TEX signature for patient prognosis. Additionally, the nomogram results showed that it outperformed other clinical characteristic models.

The TME consists of dynamic metabolic and immune interactions between pre-cancerous and cancerous tumor cells and stromal cells (such as epithelial cells, fibroblasts, endothelial cells, and hematopoietic-derived immune cells). Tumors continually adjust the TME through metabolism, ultimately forming a hypoxic, acidic, nutrient-deprived microenvironment rich in "toxic" metabolic by-products (e.g., lactic acid, kynurenine, reactive oxygen species, and adenosine)⁴⁵. In this microenvironment, infiltrating T-cells often become dysfunctional, allowing tumor cells to evade immune surveillance. We analyzed the differences in immune cell infiltration levels between the HR and LR groups, finding higher infiltration of immunosuppressive cells in the HR group and higher infiltration of tumor-promoting immune cells in the LR group. This partially explains the better prognosis in the LR group and suggests mechanisms underlying the formation of the suppressive TME in the HR group. Finally, we examined the immune subtypes between the HR and LR groups to better understand the overall immune status of the patients and the role of immune regulation in tumor progression.

ICI-based immunotherapy has become an indispensable part of treatment strategies for various cancers and is promoted as a first-line treatment for KIRC. To enable personalized and precise screening of KIRC patients for immune checkpoint blockade therapy, we evaluated the expression of immune checkpoint genes between the HR and LR groups. Studies have shown that TIGIT (T cell immunoreceptor with Ig and ITIM domains) is an inhibitory receptor primarily expressed on T-cells and natural killer (NK) cells. TIGIT signaling inhibits cytokine production and cytotoxicity of NK cells, as well as T-cell activation and proliferation, while enhancing the suppressive capacity and generation of Tregs and inducing tolerogenic dendritic cells (DCs) ^46–48. We found high expression of TIGIT in the HR group. Additionally, PDCD1, TIGIT, and HAVCR2 were significantly upregulated in the HR group in the test set. Previous studies identified these three as markers of T-cell exhaustion⁴⁹, and we observed significant upregulation of PDCD1 and TIGIT in the training set as well. The TEX scores collected in ImmuCellAI validated the accuracy of the model in predicting T-cell exhaustion scores. Furthermore, we validated the model's accuracy and effectiveness in a cohort of patients receiving immunotherapy, finding that risk scores correlated with response and prognosis to immune checkpoint therapy. This result aids clinicians in formulating precise immunotherapy regimens. We found that DERL3 is a key gene in the TEX risk score, and its high expression is associated with an increased TEX score. Single-cell analysis further revealed the possible interactions between DERL3 and TEX. Subsequent RT-PCR confirmed that DERL3 is highly expressed in KIRC patients, suggesting it could serve as a therapeutic target.

Despite the outstanding performance of our TEX signature in identifying patients' immune status and predicting their prognosis, some limitations need to be addressed in subsequent studies. First, our prognostic model was constructed and validated using retrospective data from public databases, requiring more prospective real-world data to verify its clinical utility. Second, we defined TEX based on the scores of four TEX-related signaling pathways, which may oversimplify the definition of TEX. Although basic experiments were conducted to validate several genes in the model, the specific mechanisms by which they exert their functions require further exploration.

For the first time, we comprehensively described the prognostic significance, immunotherapeutic value, and potential mechanisms of TEX in KIRC patients and developed a new TEX model identifying TEX-related prognostic genes. This model assists in probing the immune microenvironment of KIRC patients and accurately predicting their survival outcomes, aiding clinicians in identifying patient groups that can benefit most from immunotherapy and chemotherapy, aiming to achieve precision medicine.

Conflict of Interest:

The authors declare no potential conflicts of interest.

Funding:

This research was funded by National Natural Science Foundation of China, 82160474 (Y.Z.); Natural Science Foundation of Jiangxi Province, 20202BAB206018 (Y.Z.).

Author Contribution

C.Z., J.H., and Y.Z. conceived and designed the study. C.Z. and Z.R. coordinated and performed the computational analysis, including processing and downstream analysis. C.Z. and J.H. drafted the manuscript and collectively designed the figures. Y.Z. supervised the project and provided critical feedback and intellectual content to the manuscript. All authors read the manuscript and provided feedback. The work reported in the paper has been performed by the authors unless clearly specified in the text.

Data Availability

The expression data and clinical information of KIRC patients were obtained from the TCGA dataset integrated in the Xena database(https://xena.ucsc.edu/). Data from IMvigor2108 (http://research-pub.gene.com/IMvigor210CoreBiologies/) were used to assess and predict the response of TEX features to tumor immunotherapy. Single-cell sequencing data (GSE159115) of 6 benign human kidney and 7 renal tumor specimens were obtained from the GEO database(https://www.ncbi.nlm.nih.gov/geo/). TEX signaling pathway signal transduction, and marker genomes in the Molecular Signatures Database(MSigDB, V7.2) (https://www.gsea-msigdb.org/gsea/msigdb). TEX-related data on KIRC immune cell infiltration were collected from the ImmuCellAI online database (http://bioinfo.life.hust.edu.cn/).

Sung, H. et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA: a cancer journal for clinicians 71, 209-249, doi:10.3322/caac.21660 (2021).
Lai, Y. et al. The tumour microenvironment and metabolism in renal cell carcinoma targeted or immune therapy. Journal of cellular physiology 236, 1616-1627, doi:10.1002/jcp.29969 (2021).
Wang, Y. et al. Evolution of cell therapy for renal cell carcinoma. Molecular cancer 23, 8, doi:10.1186/s12943-023-01911-x (2024).
McLane, L. M., Abdel-Hakeem, M. S. & Wherry, E. J. CD8 T Cell Exhaustion During Chronic Viral Infection and Cancer. Annual review of immunology 37, 457-495, doi:10.1146/annurev-immunol-041015-055318 (2019).
Chow, A., Perica, K., Klebanoff, C. A. & Wolchok, J. D. Clinical implications of T cell exhaustion for cancer immunotherapy. Nature reviews. Clinical oncology 19, 775-790, doi:10.1038/s41571-022-00689-z (2022).
Mier, J. W. The tumor microenvironment in renal cell cancer. Current opinion in oncology 31, 194-199, doi:10.1097/cco.0000000000000512 (2019).
Li, R. et al. Mapping single-cell transcriptomes in the intra-tumoral and associated territories of kidney cancer. Cancer cell 40, 1583-1599.e1510, doi:10.1016/j.ccell.2022.11.001 (2022).
Mariathasan, S. et al. TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature 554, 544-548, doi:10.1038/nature25501 (2018).
Liberzon, A. et al. Molecular signatures database (MSigDB) 3.0. Bioinformatics (Oxford, England) 27, 1739-1740, doi:10.1093/bioinformatics/btr260 (2011).
Hänzelmann, S., Castelo, R. & Guinney, J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC bioinformatics 14, 7, doi:10.1186/1471-2105-14-7 (2013).
Miao, Y. R. et al. ImmuCellAI: A Unique Method for Comprehensive T-Cell Subsets Abundance Prediction and its Application in Cancer Immunotherapy. Advanced science (Weinheim, Baden-Wurttemberg, Germany) 7, 1902880, doi:10.1002/advs.201902880 (2020).
Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC bioinformatics 9, 559, doi:10.1186/1471-2105-9-559 (2008).
Sun, D. et al. TISCH: a comprehensive web resource enabling interactive single-cell transcriptome visualization of tumor microenvironment. Nucleic acids research 49, D1420-d1430, doi:10.1093/nar/gkaa1020 (2021).
Zhang, Z. et al. Pan-cancer landscape of T-cell exhaustion heterogeneity within the tumor microenvironment revealed a progressive roadmap of hierarchical dysfunction associated with prognosis and therapeutic efficacy. EBioMedicine 83, 104207, doi:10.1016/j.ebiom.2022.104207 (2022).
Gibbs, D. L. J. b. Robust classification of Immune Subtypes in Cancer. (2020).
Thorsson, V. et al. The Immune Landscape of Cancer. Immunity 48, 812-830.e814, doi:10.1016/j.immuni.2018.03.023 (2018).
Xu, W. X. et al. An Integrative Pan-Cancer Analysis Revealing LCN2 as an Oncogenic Immune Protein in Tumor Microenvironment. Frontiers in oncology 10, 605097, doi:10.3389/fonc.2020.605097 (2020).
Mariathasan, S. et al. TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature 554, 544-548, doi:10.1038/nature25501 (2018).
Wherry, E. J. T cell exhaustion. Nature immunology 12, 492-499, doi:10.1038/ni.2035 (2011).
Waldmann, T. The contrasting roles of IL-2 and IL-15 in the life and death of lymphocytes: implications for the immunotherapy of rheumatological diseases. Arthritis research 4 Suppl 3, S161-167, doi:10.1186/ar584 (2002).
Lodolce, J. et al. Interleukin-15 and the regulation of lymphoid homeostasis. Molecular immunology 39, 537-544, doi:10.1016/s0161-5890(02)00211-0 (2002).
Becknell, B. & Caligiuri, M. A. Interleukin-2, interleukin-15, and their roles in human natural killer cells. Advances in immunology 86, 209-239, doi:10.1016/s0065-2776(04)86006-1 (2005).
Marra, P. et al. IL15RA Drives Antagonistic Mechanisms of Cancer Development and Immune Control in Lymphocyte-Enriched Triple-Negative Breast Cancers. Cancer Research 74, 4908-4921, doi:10.1158/0008-5472.CAN-14-0637 (2014).
Vestal, D. J. & Jeyaratnam, J. A. The guanylate-binding proteins: emerging insights into the biochemical properties and functions of this family of large interferon-induced guanosine triphosphatase. Journal of interferon & cytokine research : the official journal of the International Society for Interferon and Cytokine Research 31, 89-97, doi:10.1089/jir.2010.0102 (2011).
Praefcke, G. J. K. Regulation of innate immune functions by guanylate-binding proteins. International journal of medical microbiology : IJMM 308, 237-245, doi:10.1016/j.ijmm.2017.10.013 (2018).
Chiu, H. W. et al. Guanylate binding protein 5 triggers NF-κB activation to foster radioresistance, metastatic progression and PD-L1 expression in oral squamous cell carcinoma. Clinical immunology (Orlando, Fla.) 259, 109892, doi:10.1016/j.clim.2024.109892 (2024).
Yu, X. et al. GBP5 drives malignancy of glioblastoma via the Src/ERK1/2/MMP3 pathway. Cell death & disease 12, 203, doi:10.1038/s41419-021-03492-3 (2021).
Hachim, M. Y., Hachim, I. Y., Talaat, I. M., Yakout, N. M. & Hamoudi, R. M1 Polarization Markers Are Upregulated in Basal-Like Breast Cancer Molecular Subtype and Associated With Favorable Patient Outcome. Frontiers in immunology 11, 560074, doi:10.3389/fimmu.2020.560074 (2020).
Wang, Q. et al. Distinct prognostic value of mRNA expression of guanylate-binding protein genes in skin cutaneous melanoma. Oncology letters 15, 7914-7922, doi:10.3892/ol.2018.8306 (2018).
Cheng, S. W. et al. GBP5 Repression Suppresses the Metastatic Potential and PD-L1 Expression in Triple-Negative Breast Cancer. Biomedicines 9, doi:10.3390/biomedicines9040371 (2021).
Li, Y. et al. DERL3 functions as a tumor suppressor in gastric cancer. Comput Biol Chem 84, 107172, doi:10.1016/j.compbiolchem.2019.107172 (2020).
Yu, F. et al. DERL3 suppresses colorectal cancer metastasis through negatively regulating MYCN level. Minerva medica 114, 316-322, doi:10.23736/s0026-4806.20.06657-4 (2023).
Zhou, D. et al. [Mechanism of DERL3 Affecting the Proliferation, Invasion and Metastasis of Lung Adenocarcinoma A549 Cells]. Zhongguo fei ai za zhi = Chinese journal of lung cancer 23, 638-645, doi:10.3779/j.issn.1009-3419.2020.104.22 (2020).
Lin, L. et al. Integrated profiling of endoplasmic reticulum stress-related DERL3 in the prognostic and immune features of lung adenocarcinoma. Frontiers in immunology 13, 906420, doi:10.3389/fimmu.2022.906420 (2022).
Shibata, M. et al. Overexpression of Derlin 3 is associated with malignant phenotype of breast cancer cells. Oncol Rep 38, 1760-1766, doi:10.3892/or.2017.5800 (2017).
Garrett-Sinha, L. A. et al. PU.1 and Spi-B are required for normal B cell receptor-mediated signal transduction. Immunity 10, 399-408, doi:10.1016/s1074-7613(00)80040-0 (1999).
Zhao, X. et al. SPIB acts as a tumor suppressor by activating the NFkB and JNK signaling pathways through MAP4K1 in colorectal cancer cells. Cellular signalling 88, 110148, doi:10.1016/j.cellsig.2021.110148 (2021).
Huang, Q. et al. Spi-B Promotes the Recruitment of Tumor-Associated Macrophages via Enhancing CCL4 Expression in Lung Cancer. Frontiers in oncology 11, 659131, doi:10.3389/fonc.2021.659131 (2021).
Gu, W., Qi, G. & Chen, L. SPIB Knockdown Inhibits the Immune Escape of Ovarian Cancer Cells by Reducing PD-L1 (CD274) Expression and Inactivating the JAK/STAT Pathway. Iranian journal of immunology : IJI 20, 335-347, doi:10.22034/iji.2023.98236.2559 (2023).
Jacquemin, E., Hagenbuch, B., Stieger, B., Wolkoff, A. W. & Meier, P. J. Expression cloning of a rat liver Na(+)-independent organic anion transporter. Proceedings of the National Academy of Sciences of the United States of America 91, 133-137, doi:10.1073/pnas.91.1.133 (1994).
Kullak-Ublick, G. A. et al. Molecular and functional characterization of an organic anion transporting polypeptide cloned from human liver. Gastroenterology 109, 1274-1282, doi:10.1016/0016-5085(95)90588-x (1995).
König, J., Cui, Y., Nies, A. T. & Keppler, D. Localization and genomic organization of a new hepatocellular organic anion transporting polypeptide. The Journal of biological chemistry 275, 23161-23168, doi:10.1074/jbc.M001448200 (2000).
Wlcek, K. et al. The analysis of organic anion transporting polypeptide (OATP) mRNA and protein patterns in primary and metastatic liver cancer. Cancer biology & therapy 11, 801-811, doi:10.4161/cbt.11.9.15176 (2011).
Olszewski-Hamilton, U. et al. Organic Anion Transporting Polypeptide 5A1 (OATP5A1) in Small Cell Lung Cancer (SCLC) Cells: Possible Involvement in Chemoresistance to Satraplatin. Biomarkers in cancer 3, 31-40, doi:10.4137/bic.S7151 (2011).
Riera-Domingo, C. et al. Immunity, Hypoxia, and Metabolism-the Ménage à Trois of Cancer: Implications for Immunotherapy. Physiological reviews 100, 1-102, doi:10.1152/physrev.00018.2019 (2020).
Martinet, L. & Smyth, M. J. Balancing natural killer cell activation through paired receptors. Nature reviews. Immunology 15, 243-254, doi:10.1038/nri3799 (2015).
Yu, X. et al. The surface protein TIGIT suppresses T cell activation by promoting the generation of mature immunoregulatory dendritic cells. Nature immunology 10, 48-57, doi:10.1038/ni.1674 (2009).
Stanietsky, N. et al. The interaction of TIGIT with PVR and PVRL2 inhibits human NK cell cytotoxicity. Proceedings of the National Academy of Sciences of the United States of America 106, 17858-17863, doi:10.1073/pnas.0903474106 (2009).
Krishna, C. et al. Single-cell sequencing links multiregional immune landscapes and tissue-resident T cells in ccRCC to tumor topology and therapy efficacy. Cancer cell 39, 662-677.e666, doi:10.1016/j.ccell.2021.03.007 (2021).

No competing interests reported.

Download PDF

Reviews received at journal
01 Sep, 2024
Reviewers agreed at journal
01 Sep, 2024
Reviewers agreed at journal
30 Aug, 2024
Reviewers invited by journal
30 Aug, 2024
Editor assigned by journal
30 Aug, 2024
Editor invited by journal
06 Aug, 2024
Submission checks completed at journal
02 Aug, 2024
First submitted to journal
25 Jul, 2024

You are reading this latest preprint version

T-Cell Exhaustion-Related Gene Signatures Predict Clinical Outcomes and Immunotherapy Response in Kidney Renal Clear Cell Carcinoma

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

DERL3_F:CTACTTCCTGGAGGACGTCTTC

DERL3_R:TCTGCTGTGTCCAGATGGTCAG

Results

Discussion

Conclusion

Declarations

Conflict of Interest:

Funding:

Author Contribution

Data Availability

References

Additional Declarations

Status:

Version 1