Identification of a novel anoikis-related gene signature and molecular subtype to predict prognosis and tumor microenvironment in clear cell renal cell carcinoma

doi:10.21203/rs.3.rs-3436628/v1

Download PDF

Research Article

Identification of a novel anoikis-related gene signature and molecular subtype to predict prognosis and tumor microenvironment in clear cell renal cell carcinoma

https://doi.org/10.21203/rs.3.rs-3436628/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Anoikis, an alternative form of programmed cell death, plays a pivotal role in cancer invasion and metastasis, preventing the detached cancer cells from readhering to other substrates for abnormal proliferation. However, the mechanism of anoikis in clear cell renal cell carcinoma (ccRCC) remains unknown.

Methods

ARGs(anoikis-related gene) were selected from The Cancer Genome Atlas (TCGA) database and Genecards dataset using differential expression analysis. We used an unsupervised consensus clustering algorithm to classify ccRCC patients. Gene set enrichment analysis (GSVA) and single sample gene set enrichment analysis (ssGSEA) were utilized to investigate the molecular mechanism of patients in the different subgroup. The signature incorporating ARGs was identified using univariate Cox regression analysis and LASSO regression analysis. Furthermore, a nomogram containing the signature and clinical information was developed through univariate and multivariate Cox regression analysis. Kaplan– Meier survival analysis and receiver operating characteristic (ROC) curves were applied to evaluate the predictive validity of these risk models. Finally, CIBERSOT, ESTIMATE and drug sensitivity analysis were also conducted.

Results

Our results showed that the TCGA cohorts could be divided into three subgroups which we named Group A, Group B and Group C, with a remarkable difference in immune infiltration landscape and prognosis. A fresh risk model was constructed based on the 5 prognostic ARGs (BIRC5, EDA2R, PLG, OCLN and SLPI). Kaplan-Meier survival analysis showed that the overall surviva(OS) rate of patients with low risk score was significantly higher than that of patients with high risk score. Moreover, the prognostic risk model effectively predicted OS, which was validated using train datasets. The nomogram results illustrated that the prognostic risk model was an independent prognostic predictor that distinguished it from other clinical characteristics. The CIBERSORT and ESTIMATE results illustrated a significant gap in immune infiltration landscape of patients in the low- and high-risk group. TIDE score showed a more promising immunotherapy response of ccRCC patients in low risk groups. Our drug sensitivity analysis data showed significant differences in sensitivity to different chemotherapy agents by risk group.

Conclusion

In this study, we identified anoikis-related subgroups and prognostic genes in ccRCC and integrated multiple ARGs to establish a risk-predictive model, which could be significant for understanding the molecular mechanisms and treatment of ccRCC.

anoikis

TCGA

prognostic risk model

immune infiltration

clear cell renal cell carcinoma

Renal cell carcinoma (RCC) is one of the most common malignancies of the urinary system, according to the Global Cancer Statistics in 2020, the incidence and mortality of kidney cancer were 431,288 and 179,368, respectively[1]. Clear cell renal cell carcinoma (ccRCC), the most common histological subtype, is the leading cause of death of RCC patients [2]. About 20-30% of patients are metastatic at the early stages of diagnosis, 30-50% progress to metastases with local disease, and nearly 40% of patients with local RCC have distant metastases even after surgery[3]. Moreover, ccRCC is chemo- and radio-resistant neoplasia and alternative treatment options have been limited [4]. In recent years, targeted therapies and immunotherapies have further improved the prognosis of ccRCC. However, only a small percentage of ccRCC patients can benefit from these therapies[5, 6]. Therefore, there is an urgent need to identify new biomarkers to predict patient survival and response to targeted therapies and immunotherapies.

Anoikis is caused by loss of cell adhesion to the cell-extracellularmatrix (ECM) or abnormal cell adhesion apoptosis, a special form of programmed apoptosis, can effectively remove misaligned or shed cells and help maintain tissue homeostasis[7]. It is mainly triggered by the interaction of two apoptotic pathways, interference with mitochondria or activation of death receptors on the cell surface, was first identified in epithelial and endothelial cells and has been confirmed to be an important mechanism of cancer invasion and metastasis[8]. The generation of anoikis resistance can help isolated cells bypass death signaling pathways, allowing the cells to survive under adverse conditions[9]. Anoikis has therefore been identified as a vital factor in tumor progression[10].

In recent years, an increasing number of anoikis related genes (ARGs) have been identified to play important roles in various tumors. Jin et al. confirmed that GDH1-mediated metabolic reprogramming of glutaminolysis has been verified to promote anoikis resistance and tumor metastasis in LKB1-deficient lung cancer[11]. In another study by Yin Yu et al., they found that CircCEMIP promotes anoikis-resistance by enhancing protective autophagy in prostate cancer cells[12]. A previous study revealed that CPT1A-mediated fatty acid oxidation promotes colorectal cancer cell metastasis by inhibiting anoikis[13]. Moreover, ARGs have been proven to be associated with prognosis in several tumors such as gastric cancer, triple-negative breast cancer and so on[14,15]. However, a comprehensive analysis of the effects of anoikis in ccRCC is still scarce. Therefore, it is critical to identify dominant genes with prognostic significance associated with anoikis in ccRCC.

Identification of anoikis-related genes in ccRCC

The RNA sequencing data and corresponding clinical information of 607

ccRCC patients, including 535 ccRCC samples and 72 noncancerous adjacent samples (NATs), were extracted from the TCGA database. A total of 652 ARGs were downloaded from the GeneCards database. Then, the differential analysis of the ARGs between ccRCC and NATs was conducted using the “limma” R package after obtaining intersection from the above two datasets. The p-value was adjusted with the false discovery rate (FDR). FDR <0.05 and log (FC) ≥ 1 was considered statistically significant. A heatmap and volcano plot were generated using the “pheatmap” R package to visualize differentially expressed ARGs in ccRCC.

Identification of prognostic anoikis-related genes

Through differential analysis and prognostic analysis, 19 anoikis related genes such as NTRK2, BIRC5, E2F1 were found to be significantly related to the prognosis of ccRCC, and further network diagram was used to visualize the relationship between these anoikis related genes and the prognostic value of ccRCC. p < 0.05 is regarded as statistically significant.

Identification of subtype based on prognostic anoikis-related genes

We use the “ConsensusClusterPlus” package with K = 3 to hierarchically clustered the prognostic anoikis-related genes which we named GroupA, GroupB and GroupC. The “limma” package was used to identify the differentially expressed genes (DEGs) between three groups. Then, we use principal component analysis(PCA) to verify the validity of the classification.

Tumor immune microenvironment of different subtype

we compared the immune cell infiltration between the three groups, and the results were visualized using the violin plots. Heat maps were generated to show the correlation between the differential expression of anoikis-related genes in different subtypes and the clinical characteristics of ccRCC patients. Gene set enrichment analysis (GSEA) was performed to identify enriched pathways using relative gene sets.

Construction of a prognostic model based on anoikis-related genes

The ccRCC patients was randomly divided into the training and

testing datasets at a 1:1 ratio using the “caret” package. Univariate Cox regression, LASSO regression and multivariate Cox regression analysis were used to construct a prognostic risk score model based on the selected 5 anoikis-related genes using the training datasets. The risk score was calculated as following: Riskscore=CoefBIRC5×ExpressionBIRC5+CoefEDA2R×ExpressionEDA2R + CoefPLG×ExpressionPLG+CoefOCLN ×ExpressionOCLN + CoefSLPI ×ExpressionSLPI, where Coef represents the coefficient with the lowest Akaike’s information criterion (AIC) values. Coef >0 and Coef <0 indicate risk and protective factors, respectively. Based on the median risk score, all the ccRCC samples were assigned to the high-risk and low-risk groups. We then calculated the predictive accuracy of the prognostic model for OS using the "Time-ROC" package.

Independent prognosis analysis of risk model

Univariate and multivariate Cox regression analyses were used to assess the independence of the risk model by the R package "survival". A nomogram model of clinicopathological features and risk scores was established by the R package "rms". Based on the Cox regression analysis, all variates were calculated and evaluated the 1-, 3-, and 5-year survival probability of ccRCC. Subsequently, the decision curve analysis (DCA) was used to evaluate the clinical usefulness of the nomogram.

Immune microenvironment landscape and drug sensitivity analysis

The CIBERSORT algorithm was used to estimate the score of 22-type immune cells based on "CIBERSORT R script v1.03". The results are graphically represented. Then the stromal, immune, and ESTIMATE scores of ccRCC were estimated using the R package “estimate”. Tumor Immune Dysfunction and Exclusion (TIDE) scores of high- and low-risk group were analyzed use the TIDE database. Finally, based on the Genomics of Cancer Drug Sensitivity (GDSC) database, the R package "pRRophetic" was used to predict the antineoplastic drug response of each ccRCC sample in low risk and high risk.

Statistical analysis

Statistical analyses of this study were performed using the R software (version 4.2.2). For comparison of the differences of categorical and continuous variables, the Chi-square and Student’s t-tests were utilized. DEGs were identified using the “limma” package, and hierarchical clustering analysis was performed using the “ConsensusClusterPlus” package. Univariate Cox regression, LASSO regression and multivariate Cox regression analyses were performed to identify the independent prognostic risk factors, which were used for the construction of a anoikis-related prognostic signature. All statistical analyses were visualized using the "ggplot 2" R package. For each analysis in the present study, statistical significance was set at p < 0.05.

Identification of anoikis-related genes in ccRCC

The flowchart of this study is shown in Figure1. We obtained RNA transcriptome data and corresponding clinical data of 607 ccRCC patients in TCGA database. Among 652 ARGs identified from the GeneCards database, 32 differentially expressed ARGs were significant between tumor and noncancerous adjacent samples (NATs), including 17 upregulated and 15 downregulated in tumors (FIGURE 2A, B).

The biological roles of anoikis-related genes in different subgroups of ccRCC and functional analysis

We identified 19 prognostic genes using the univariate Cox regression analysis(FIGURE 3A). Regulatory relationships among prognostic genes is shown in FIGURE 3B. The ccRCC cohort was divided into three distinct groups, which we named as GroupA, GroupB and GroupC, based on these prognostic genes using the “ConsensusClusterPlus” package with K = 3(FIGURE 3C). The PCA and tSNE score plot illustrated a remarkable separation between the three subgroups(FIGURE 3E, F). As shown in Figure 3D, patients in Group A were significantly associated with a better prognosis while in Group B have a worst prognosis(p < 0.001). The relationship between the expressions of prognostic genes and clinicopathological parameters such as age, gender, TNM stage(T stage, N stage, and M stage) was shown using the Heatmap(FIGURE 4B). As we can see, the gene PTHLH,UBE2C, MMP9 and BIRC5 are high expression in Group B but low expression in Group A and Group B(FIGURE 4A). On the contrary, the gene PLG and OCLN are high expression in Group A but low expression in Group B and Group C.

Immune cell infiltration in the three groups was analyzed. As shown in Figures 4C, Activated CD4 T cell and Natural killer T cell were predominant in Group B, while Type 2 T helper cell Tcells were predominant in Group A (p < 0.05).

To explore the underlying mechanisms and functions, functional enrichment analysis was performed to reveal the potential biological functions associated with ARGs involved in typing. GSVA analysis shows that the Group A were mainly clustered in several termsm, such as vibrio cholerae infection, aldosterone regulated sodium reabsorption, tight junction, fatty acid metabolism and so on. Inversely, the worst prognosis Group B were mainly clustered in jak stat signaling pathway, cytokine-cytokine receptor interaction, toll like receptor signaling pathway and so on(FIGURE 5A, B, C). At the same time, GSEA analysis shows that Group B were significantly enriched in several biological processes or molecular functions, such as chemokine signaling pathway, hematopoietic cell lineage, primary immunodeficiency and T cell receptor signaling pathway(FIGURE 5 D, E, F). In conclusion, these results suggested that these signaling pathways be mainly associated with poor prognosis in ccRCC patients.

Development and validation of the anoikis-related genes signature

To establish the signature for predicting the survival status of ccRCC patients, a total of 518 patients, who met the inclusion and exclusion criteria, were randomly grouped into a training set (259 patients) and a testing set (259 patients) at a 1:1 ratio. In the training set, we performed univariate Cox regression on the differentially expressed ARGs and dentified 19 genes associated with prognosis. Five optimal prognostic ARGs were further selected by multivariate Cox regression analysis and LASSO regression analysis (FIGURE 6A, B). Finally, a risk score model including the selected ARGs was developed in the training set: Risk score=coefficients*Expression of ARGs(i)(Table 1). Based on the median risk score, all the patients were divided into the high- and low-risk groups. Then, We drew a heatmap according to the expression of prognostic genes in the high and low risk groups. As shown in Figures 7D, BIRC5 and SLPI are high expression in high risk groups while EDA2R, PLG and SLPI are low expression in high risk groups. The relationship between the cluster classification of ccRCC patients and the risk score is shown in Figure 7E, and the risk score of Group B is higher than that of the other two groups. In the Sankey plot (FIGURE 7F), further delineation of subgroups with respect to risk-score groups and vital status revealed that most patients in Group B had a high-risk score and a worse prognosis. Survival curve (Kaplan-Meier) analysis results showed that the OS for patients in the low-risk group was significantly better than that of patients in the high-risk group, in both the training and testing datasets (p < 0.001)(FIGURE 6C). To evaluate the predictive accuracy of our model, we plotted the ROC curves for both the training and testing datasets(FIGURE 6E, F). The AUC values for OS at 1 years, 3 years, and 5 years in both datasets were >0.5, indicating that the prognostic model could predict the survival of ccRCC patients with considerable accuracy. Univariate and multivariate Cox regression analyses were used to evaluate the risk score based on the ARGs as an independent prognosis predictor for ccRCC. Univariate Cox regression analysis showed that age (hazard ratio (HR) = 1.023, p = 0.011), stage T (HR = 1.934, p < 0.001), stage N (HR = 3.422, p < 0.001), stage M (HR = 4.326, p < 0.001) and risk score (HR = 1.139, p < 0.001) were closely correlated with OS rate in ccRCC(FIGURE 7A). Multivariate Cox regression analysis result indicated that age (HR = 1.034, p < 0.001), stage T (HR = 1.369, p = 0.019), stage M (HR = 3.181, p < 0.001) and risk score (HR = 1.119, p < 0.001) were an independent prognosis indicator for ccRCC(FIGURE 7B, C).

Construction and evaluation of the nomogram incorporating the anoikis-related gene signature

All variables which were significant in the multivariate Cox regression analysis were enrolled in the predictive model to predict 1-year, 3-year and 5-year OS for ccRCC patients(FIGURE 8A). The calibration curve showed that the 1-, 3-, and 5-year overall survival rates predicted by the nomogram were basically consistent with the actual overall survival rates(FIGURE 8B). According to the total points generated from the nomogram, ccRCC patients in the entire datasets were grouped into different risk groups by the median value of the total points. Figure 8C demonstrated that patients in the high-points group possessed higher cumulative hazard compared to those in the low-points group. The DCA analysis illustrated that a higher net benefit could be obtained through using the nomogram to reach a decision than using other factors(FIGURE 8D-F).

Differential analysis of immune cells and tumor microenvironment

In order to explore the differences of immune cells, we performed the CIBERSORT algorithm to explore the correlation between infiltrating immune cells and risk group. The results showed that the infiltration of immune cells was different between the two risk groups, and the infiltration of immune cells such as memory B cells and activated memory T cells CD4 were higher in the high-risk group while resting dendritic cells and resting mast cells were higher in the low-risk group, the difference was statistically significant (P < 0.05)(FIGURE 9A, B). The correlation between the risk score and tumor infiltrating immune cells is shown in Figure 4C. The risk score is significantly positively correlated with the infiltration degree of T cells follicular helper, activated memory T cells CD4 and macrophages M0 but significantly negatively correlated with resting mast cells and resting Dendritic cells (P < 0.001)(FIGURE 10A-L). In addition, TIDE analysis showed that the high-risk group had a higher TIDE score than the low-risk group(FIGURE 9D).

Meanwhile, we plotted the violin plot of the tumor stromal score, immune score and ESTIMATE score of the high and low risk groups, and found that the high-risk group had higher tumor stroma score, immune score and ESTIMATE score(FIGURE 9E).

Differential analysis of drug sensitivity

According to drug sensitivity analysis, nine cancer drugs were screened from the GDSC database, and their drug sensitivity was significantly different between the high and low risk groups(P < 0.001). The results in Figure 11A-I show that the high-risk group showed high sensitivity to the drugs 5−Fluorouracil, Bortezomib, Dactolisib, Paclitaxel, Vinorelbine and Vinblastine.

Based on the differential expression of anoikis-related genes in ccRCC tumor tissues and normal tissues, we analyzed the effects of anoikis-related genes on the tumor immune microenvironment and constructed a prognostic risk model. We used the unsupervised consensus clustering algorithm to classify ccRCC patients and obtained three potential clusters GroupA, GroupB and GroupC. The analysis and evaluation results showed that there was a significant difference in overall survival of ccRCC patients between the three clusters, with Group B having a lowing probability of survival. We used GSVA and ssGSEA to analyze the biological processes and infiltrating immune cells in these subgroups and found that Group B was enriched in cytokine-cytokine receptor interaction, hematopoietic cell lineage, primary immunodeficiency, jak stat signaling pathway, toll like receptor signaling pathway, chemokine signaling pathway and T cell receptor signaling pathway. This suggests that these biological processes and signaling pathways may enhance the invasiveness of renal clear cell tumors and increase the probability of tumor metastasis. Then we screened five genes highly correlated with HCC and constructed a prognostic risk model. Finally, We further evaluated the immune infiltration landscape involved and drug sensitivity analyses.

The BIRC5 (Baculoviral IAP Repeat Containing 5) gene ocated on chromosome 17 (17q25.3) encodes for the Survivin protein, which is a member of the IAPs (inhibitor of apoptosis family) that is normally expressed in humans during fetal development and in adult proliferating cells[16]. Previous studies have shown that BIRC5/Survivin is a dual cellular functions protein that directly regulates both apoptosis and mitosis in embryonic cells during embryogenesis and in cancer cells during tumorigenesis and tumor metastasis[17]. Survivin is a small protein with different isoforms, the majority of which are related to inhibition of apoptosis and promotion of cell proliferation[18]. Some studies indicate that high BIRC5 expression was significantly associated with worse OS for breast-, hepatocellular-, and gastric cancer[19, 20, 21]. Our study showed that BIRC5 was highly expressed in ccRCC tissues and was closely associated with poor prognosis of patients in the high-risk group. This may be related to inhibiting the anoikis and promoting the proliferation of tumor cells in ccRCC.

Secretory Leukocyte Protease Inhibitor (SLPI) is an evolutionarily conserved, pleiotropic protein expressed at mucosal surfaces, mainly by epithelial cells[22]. SLPI maintains homeostasis at barrier tissues by preventing tissue destruction and regulating the threshold of inflammatory immune responses, while protecting the host from infection[23]. However, excessive expression of SLPI in cancer cells may have detrimental consequences, as recent studies demonstrate that overexpression of SLPI increases the metastatic potential of epithelial tumors[24]. SLPI is upregulated in several types of carcinoma. High SLPI mRNA expression correlates with shorter overall survival and shorter distant metastasis-free survival in patients with triple-negative breast cancer[25]. SLPI expression is increased in ovarian cancer compared to normal ovaries[26]. Patients with non-small cell lung cancer also have higher serum SLPI levels compared to healthy controls[27]. The role of SLPI in cancer has not been fully elucidated, and its role may be related to promoting tumor cells to form vessel-like structures, promoting cell invasion by inducing MMP-2 and MMP-9, and inhibiting cell apoptosis by inducing cyclin D1[24, 29, 30]. Research has shown that SLPI is hypomethylated and overexpressed in kidney renal papillary cell carcinoma and associated with tumor progression and poor prognosis[31]. However, the role of SLPI in ccRCC has not been studied before, and our study is the first to find that SLPI is highly expressed in ccRCC and associated with worse prognosis.

EDA2R, also known as XEDAR (X-linked Ectodermal Dysplasia Receptor), encodes a transmembranal receptor that belongs to the tumor necrosis factor (TNF)-receptor superfamily[32]. The study by Tanikawa et al.identified EDA2R as a p53 target gene that mediates anoikis, and suggested that EDA2R is a tumor suppressor that limits colorectal cancer progression[33]. Their study further found that XEDAR is a putative tumor suppressor that could prevent malignant transformation and tumor progression by regulating apoptosis and anoikis[33]. In another study it was found that XEDAR expression is downregulated in most breast cancers via promoter methylation, which may contribute to accelerated tumor development by blocking EDA-A2-induced cell death[34]. XEDAR may represent a novel breast tumor suppressor gene, and restoration of its expression by treatment with DNA demethylating agents may represent an attractive approach for the treatment of breast cancer[34]. Our research supports this view. Our study found that EDA2R is highly expressed in low-risk ccRCC patients. This suggests that EDA2R may inhibits invasion and metastasis of renal clear cell tumors.

PLG plays an important role in tissue remodeling during development, physical injury, inflammation and carcinogenesis[35]. It can help degrade the extracellular matrix with other matrix metalloproteases, such as collagenases, gelatinases and stromelysins, which all serve a vital character in cancer invasion, especially in lung and breast cancer[36]. However, PLG is not only a pro-tumorigenic factor but also an anti-tumorigenic factor due to the fact that proteolysis of PLG can release angiotensin, which will function against cancer progression[37]. This may explain the result that the expression of PLG in high-risk ccRCC samples was lower than that in low-risk ccRCC samples in our study, which meant the low expression of PLG was important for ccRCC progression. In addition to our results, down-regulation of PLG in ccRCC was confirmed by Schröd-ter et al. who screened the DEGs using a microarray and qPCR[38]. Worse OS of ccRCC patients associated with low expression of PLG was verified by Wang et al.[39].

The OCLN gene is responsible for producing the occludin protein, which is present in all tight junctions[40]. Tight junctions determine epithelial cell polarity and undergo dissolution during cancer progression that is accompanied by cytoskeletal changes[41]. Downregulation of OCLN has been observed in liver tumors, breast cancer, endometrial and lung carcinomas[42-45]. A study of human clear cell renal cell carcinoma observed that the silencing of OCLN and GAS1 may contribute to tumor progression due to a possible relationship between these genes and cell-cell adhesion, apoptosis and proliferation[46]. Another study also proved that the expression of OCLN was downregulated in ccRCC cells and overexpression of OCLN could reduce the proliferation, migration, and invasion of ccRCC cells[47]. In the present study, we observed that the expression of PLG in ccRCC samples was lower than that in normal tissue and low expression of OCLN is associated with worse prognosis.

ccRCC is considered to be an immunogenic tumor, and a large number of immune cells, such as tumor-infiltrating lymphocytes, can be detected in the tumor tissue. Therefore, the use of immunotherapy to generate an effective immune response against the tumor to delay the development of cancer is thought to be effective in ccRCC[48]. According to our immune infiltration analysis results, high-risk ccRCC had an immune microenvironment consisting of higher levels of memory B cells and activated memory T cells CD4, and lower resting dendritic cells and resting mast cells. It is also suggested that immunotherapy is more beneficial in high-risk RCC patients. A previous study indicate that a pre-existing high number of memory B cells is associated with higher probability of response to nivolumab and longer PFS and OS in patients with m-ccRCC[49]. Therefore, our findings support that patients in the high-risk group may benefit more from immunotherapy.

The tumor microenvironment includes infiltrating immune cells and extracellular matrix. Extracellular matrix (ECM) plays an important role in tumor development, including induction of anoikis resistance and cell adhesion mediated drug resistance[50]. In our study, we found that high-risk ccRCC patients had significantly higher stromal scores than low-risk patients. It indicates that extracellular matrix plays an important role in the development and metastasis of ccRCC. On the other hand, TIDE analysis showed that the high-risk group had a higher TIDE score than the low-risk group. Higher tumor TIDE prediction scores were associated not only with poorer immune checkpoint inhibition therapy but also with poorer patient survival under anti-PD1 and anti-CTLA4 therapy[51]. This indicates that patients in the high-risk group have a higher risk of immune escape and resistance to immunotherapy. Although there is evidence that the use of immunotherapy to generate an effective immune response against the tumor to delay the development of cancer is considered to be effective in ccRCC. However, more than half of the patients could not achieve the long-term response to PD-1-related treatment[52]. Therefore, it is important to further screen ccRCC patients suitable for immunotherapy. Additionally, our drug sensitivity analysis data showed significant differences in susceptibility to different chemotherapy agents in different risk strata after risk modeling using anoikis related genes. This provides new insights into precisely targeted therapy for ccRCC patients.

In the present study, We established a prognostic anoikis-related gene signature model for ccRCC, along with a predictive nomogram based on the 5-gene risk score, clinical stages and age of the patient. Then we preliminarily evaluated the efficacy of risk model in predicting the prognosis of ccRCC patients. Based on the differential genes, we classified UCEC patients and obtained three potential clusters (A, B and C). The analysis and evaluation results showed that there were significant differences in overall survival, biological function and immune microenvironment of UCEC patients among the three clusters. This suggests that our classification can provide more accurate diagnosis and treatment options for ccRCC patients. In addition, We preliminarily revealed the link between this risk model and RCC tumor microenvironment by immune infiltration landscape analysis and tumor microenvironment analysis. In the end, when analyzing treatment outcomes based on risk scores in ccRCC patients, we found an interaction between drug sensitivity and risk scores. However, this study still has some shortcomings. The results based on bioinformatics in this paper have not been verified by in vitro experiments. In the next step, we will experimentally verify the significance of anoikis related genes in ccRCC. In conclusion, our findings provide new insights and perspectives for novel potential therapeutic strategies and immunotherapy targets for ccRCC.

ccRCC: Clear cell renal cell carcinoma

RCC: Renal cell carcinoma

ARGs: Anoikis-related gene

GSVA: Gene set enrichment analysis

ssGSEA: Single sample gene set enrichment analysis

LASSO: Least absolute shrinkage and selection operator

ROC: Receiver operating characteristic

OS: Overall surviva

ECM: Cell-extracellularmatrix

NATs: Noncancerous adjacent samples

DEGs: Differentially expressed genes

PCA: Principal component analysis

tSNE: t-Distributed Stochastic Neighbor Embedding

GSEA: Gene set enrichment analysis

AIC: Akaike’s information criterion

DCA: Decision curve analysis

TIDE: Tumor Immune Dysfunction and Exclusion

GDSC: Genomics of Cancer Drug Sensitivity

BIRC5: Baculoviral IAP Repeat Containing 5

SLPI: Secretory Leukocyte Protease Inhibitor

XEDAR: X-linked Ectodermal Dysplasia Receptor

TNF: Tumor necrosis factor

PD-L1: Programmed death-ligand 1

Acknowledgements

Not applicable.

Authors’ contributions

Shaoquan Li, designed the study, wrote the manuscript, and collected the corresponding datasets. Peng Luo and Shusheng Yuan assisted in the model construction and model validation. Shuibo Shi drew the figures and tables, Weimin Chen supervised the whole project and edited the manuscript. All authors read and approved the final manuscript.

Funding

Not applicable.

Availability of data and materials

This study analyzed publicly available datasets. These datasets are available from the TCGA .

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests

Bahadoram S, Davoodi M, Hassanzadeh S, Bahadoram M, Barahman M, Mafakher L. Renal cell carcinoma: an overview of the epidemiology, diagnosis, and treatment. G Ital Nefrol. 2022 Jun 20;39(3):2022-vol3.
Hsieh JJ, Purdue MP, Signoretti S, Swanton C, Albiges L, Schmidinger M, Heng DY, Larkin J, Ficarra V. Renal cell carcinoma. Nat Rev Dis Primers. 2017 Mar 9;3:17009.
Klatte T, Rossi SH, Stewart GD. Prognostic factors and prognostic models for renal cell carcinoma: a literature review. World J Urol. 2018 Dec;36(12):1943-1952.
Geissler K, Fornara P, Lautenschläger C, Holzhausen HJ, Seliger B, Riemann D. Immune signature of tumor infiltrating immune cells in renal cancer. Oncoimmunology. 2015 Feb 3;4(1):e985082.
Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC, McDermott DF, Powderly JD, Carvajal RD, et al.Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N Engl J Med. 2012 Jun 28;366(26):2443-54.
Motzer RJ, Rini BI, McDermott DF, Redman BG, Kuzel TM, Harrison MR, et al. Nivolumab for Metastatic Renal Cell Carcinoma: Results of a Randomized Phase II Trial. J Clin Oncol. 2015 May 1;33(13):1430-7.
Taddei ML, Giannoni E, Fiaschi T, Chiarugi P. Anoikis: an emerging hallmark in health and diseases. J Pathol. 2012 Jan;226(2):380-93. doi: 10.1002/path.3000. PMID: 21953325.
Paoli P, Giannoni E, Chiarugi P. Anoikis molecular pathways and its role in cancer progression. Biochim Biophys Acta. 2013 Dec;1833(12):3481-3498.
Adeshakin FO, Adeshakin AO, Afolabi LO, Yan D, Zhang G, Wan X. Mechanisms for Modulating Anoikis Resistance in Cancer and the Relevance of Metabolic Reprogramming. Front Oncol. 2021 Mar 29;11:626577.
Diao X, Guo C, Li S. Identification of a novel anoikis-related gene signature to predict prognosis and tumor microenvironment in lung adenocarcinoma. Thorac Cancer. 2023 Jan;14(3):320-330.
Jin L, Chun J, Pan C, Kumar A, Zhang G, Ha Y, Li D, Alesi GN, Kang Y, et al. The PLAG1-GDH1 Axis Promotes Anoikis Resistance and Tumor Metastasis through CamKK2-AMPK Signaling in LKB1-Deficient Lung Cancer. Mol Cell. 2018 Jan 4;69(1):87-99.e7.
Yu Y, Song Y, Cheng L, Chen L, Liu B, Lu D, Li X, Li Y, Lv F, Xing Y. CircCEMIP promotes anoikis-resistance by enhancing protective autophagy in prostate cancer cells. J Exp Clin Cancer Res. 2022 Jun 2;41(1):188.
Wang YN, Zeng ZL, Lu J, Wang Y, Liu ZX, He MM, Zhao Q, Wang ZX, et al. CPT1A-mediated fatty acid oxidation promotes colorectal cancer cell metastasis by inhibiting anoikis. Oncogene. 2018 Nov;37(46):6025-6040.
Ye G, Yang Q, Lei X, Zhu X, Li F, He J, Chen H, et al. Nuclear MYH9-induced CTNNB1 transcription, targeted by staurosporin, promotes gastric cancer cell anoikis resistance and metastasis. Theranostics. 2020 Jun 12;10(17):7545-7560.
D'Amato NC, Rogers TJ, Gordon MA, Greene LI, Cochrane DR, et al. A TDO2-AhR signaling axis facilitates anoikis resistance and metastasis in triple-negative breast cancer. Cancer Res. 2015 Nov 1;75(21):4651-64.
Fäldt Beding A, Larsson P, Helou K, Einbeigi Z, Parris TZ. Pan-cancer analysis identifies BIRC5 as a prognostic biomarker. BMC Cancer. 2022 Mar 25;22(1):322.
Lin TY, Chan HH, Chen SH, Sarvagalla S, Chen PS, Coumar MS, et al. BIRC5/Survivin is a novel ATG12-ATG5 conjugate interactor and an autophagy-induced DNA damage suppressor in human cancer and mouse embryonic fibroblast cells. Autophagy. 2020 Jul;16(7):1296-1313.
Khan S, Ferguson Bennit H, Asuncion Valenzuela MM, Turay D, Diaz Osterman CJ, et al. Localization and upregulation of survivin in cancer health disparities: a clinical perspective. Biologics. 2015 Jul 9;9:57-67.
Dai JB, Zhu B, Lin WJ, Gao HY, Dai H, Zheng L, Shi WH, Chen WX. Identification of prognostic significance of BIRC5 in breast cancer using integrative bioinformatics analysis. Biosci Rep. 2020 Feb 28;40(2):BSR20193678.
Cao L, Li C, Shen S, Yan Y, Ji W, Wang J, Qian H, Jiang X, Li Z, Wu M, Zhang Y, Su C. OCT4 increases BIRC5 and CCND1 expression and promotes cancer progression in hepatocellular carcinoma. BMC Cancer. 2013 Feb 22;13:82.
Zou J, Liao X, Zhang J, Wang L. Dysregulation of miR-195-5p/-218-5p/BIRC5 axis predicts a poor prognosis in patients with gastric cancer. J Biol Regul Homeost Agents. 2019 Sep-Oct;33(5):1377-1385.
Nugteren S, Samsom JN. Secretory Leukocyte Protease Inhibitor (SLPI) in mucosal tissues: Protects against inflammation, but promotes cancer. Cytokine Growth Factor Rev. 2021 Jun;59:22-35.
Baranger K, Zani ML, Labas V, Dallet-Choisy S, Moreau T. Secretory leukocyte protease inhibitor (SLPI) is, like its homologue trappin-2 (pre-elafin), a transglutaminase substrate. PLoS One. 2011;6(6):e20976.
Wagenblast E, Soto M, Gutiérrez-Ángel S, Hartl CA, Gable AL, et al. A model of breast cancer heterogeneity reveals vascular mimicry as a driver of metastasis. Nature. 2015 Apr 16;520(7547):358-62.
Kozin SV, Maimon N, Wang R, Gupta N, Munn L, Jain RK, Garkavtsev I. Secretory leukocyte protease inhibitor (SLPI) as a potential target for inhibiting metastasis of triple-negative breast cancers. Oncotarget. 2017 Nov 26;8(65):108292-108302.
Hough CD, Cho KR, Zonderman AB, Schwartz DR, Morin PJ. Coordinately up-regulated genes in ovarian cancer. Cancer Res. 2001 May 15;61(10):3869-76.
Ameshima S, Ishizaki T, Demura Y, Imamura Y, Miyamori I, Mitsuhashi H. Increased secretory leukoprotease inhibitor in patients with nonsmall cell lung carcinoma. Cancer. 2000 Oct 1;89(7):1448-56.
Hoskins E, Rodriguez-Canales J, Hewitt SM, Elmasri W, Han J, Han S, Davidson B, Kohn EC. Paracrine SLPI secretion upregulates MMP-9 transcription and secretion in ovarian cancer cells. Gynecol Oncol. 2011 Sep;122(3):656-62.
Zhang D, Simmen RC, Michel FJ, Zhao G, Vale-Cruz D, Simmen FA. Secretory leukocyte protease inhibitor mediates proliferation of human endometrial epithelial cells by positive and negative regulation of growth-associated genes. J Biol Chem. 2002 Aug 16;277(33):29999-30009.
Wang YL, Zhang YY. cg04448376, cg24387542, cg08548498, and cg14621323 as a Novel Signature to Predict Prognosis in Kidney Renal Papillary Cell Carcinoma. Biomed Res Int. 2020 Dec 17;2020:4854390.
Yan M, Wang LC, Hymowitz SG, Schilbach S, Lee J, Goddard A, de Vos AM, Gao WQ, Dixit VM. Two-amino acid molecular switch in an epithelial morphogen that regulates binding to two distinct receptors. Science. 2000 Oct 20;290(5491):523-7.
Tanikawa C, Furukawa Y, Yoshida N, Arakawa H, Nakamura Y, Matsuda K. XEDAR as a putative colorectal tumor suppressor that mediates p53-regulated anoikis pathway. Oncogene. 2009 Aug 27;28(34):3081-92.
Punj V, Matta H, Chaudhary PM. X-linked ectodermal dysplasia receptor is downregulated in breast cancer via promoter methylation. Clin Cancer Res. 2010 Feb 15;16(4):1140-8.
Danø K, Rømer J, Nielsen BS, Bjørn S, Pyke C, Rygaard J, Lund LR. Cancer invasion and tissue remodeling--cooperation of protease systems and cell types. APMIS. 1999 Jan;107(1):120-7.
Almholt K, Juncker-Jensen A, Lærum OD, Johnsen M, Rømer J, Lund LR. Spontaneous metastasis in congenic mice with transgenic breast cancer is unaffected by plasminogen gene ablation. Clin Exp Metastasis. 2013 Mar;30(3):277-88.
Schrödter S, Braun M, Syring I, Klümper N, Deng M, Schmidt D, Perner S, Müller SC, Ellinger J. Identification of the dopamine transporter SLC6A3 as a biomarker for patients with renal cell carcinoma. Mol Cancer. 2016 Feb 2;15:10.
Wang S, Yu ZH, Chai KQ. Identification of EGFR as a Novel Key Gene in Clear Cell Renal Cell Carcinoma (ccRCC) through Bioinformatics Analysis and Meta-Analysis. Biomed Res Int. 2019 Feb 13;2019:6480865.
De Wever O, Pauwels P, De Craene B, Sabbah M, Emami S, Redeuilh G, Gespach C, Bracke M, Berx G. Molecular and pathological signatures of epithelial-mesenchymal transitions at the cancer invasion front. Histochem Cell Biol. 2008 Sep;130(3):481-94.
Thiery JP, Sleeman JP. Complex networks orchestrate epithelial-mesenchymal transitions. Nat Rev Mol Cell Biol. 2006 Feb;7(2):131-42.
Orbán E, Szabó E, Lotz G, Kupcsulik P, Páska C, Schaff Z, Kiss A. Different expression of occludin and ZO-1 in primary and metastatic liver tumors. Pathol Oncol Res. 2008 Sep;14(3):299-306.
Martin TA, Mansel RE, Jiang WG. Loss of occludin leads to the progression of human breast cancer. Int J Mol Med. 2010 Nov;26(5):723-34.
Tobioka H, Isomura H, Kokai Y, Tokunaga Y, Yamaguchi J, Sawada N. Occludin expression decreases with the progression of human endometrial carcinoma. Hum Pathol. 2004 Feb;35(2):159-64.
Tobioka H, Tokunaga Y, Isomura H, Kokai Y, Yamaguchi J, Sawada N. Expression of occludin, a tight-junction-associated protein, in human lung carcinomas. Virchows Arch. 2004 Nov;445(5):472-6.
Conceição AL, Da Silva CT, Badial RM, Valsechi MC, Stuqui B, Gonçalves JD, Jasiulionis MG, De Freitas Calmon M, Rahal P. Downregulation of OCLN and GAS1 in clear cell renal cell carcinoma. Oncol Rep. 2017 Mar;37(3):1487-1496.
Peng L, Cao Z, Wang Q, Fang L, Yan S, Xia D, Wang J, Bi L. Screening of possible biomarkers and therapeutic targets in kidney renal clear cell carcinoma: Evidence from bioinformatic analysis. Front Oncol. 2022 Oct 13;12:963483.
Kuusk T, Albiges L, Escudier B, Grivas N, Haanen J, Powles T, Bex A. Antiangiogenic therapy combined with immune checkpoint blockade in renal cancer. Angiogenesis. 2017 May;20(2):205-215.
Carril-Ajuria L, Desnoyer A, Meylan M, Dalban C, Naigeon M, Cassard L, Vano Y, et al. Baseline circulating unswitched memory B cells and B-cell related soluble factors are associated with overall survival in patients with clear cell renal cell carcinoma treated with nivolumab within the NIVOREN GETUG-AFU 26 study. J Immunother Cancer. 2022 May;10(5):e004885.
Wang L, Li C, Wang J, Yang G, Lv Y, Fu B, Jian L, Ma J, Yu J, Yang Z, Wu P, Li G, Liu X, Kang Z, Wang Z, Wang L, Wang H, Xu W. Transformable ECM Deprivation System Effectively Suppresses Renal Cell Carcinoma by Reversing Anoikis Resistance and Increasing Chemotherapy Sensitivity. Adv Mater. 2022 Oct;34(43):e2203518.
Chen YP, Lv JW, Mao YP, Li XM, Li JY, Wang YQ, Xu C, Li YQ, He QM, Yang XJ, Lei Y, et al. Unraveling tumour microenvironment heterogeneity in nasopharyngeal carcinoma identifies biologically distinct immune subtypes predicting prognosis and immunotherapy responses. Mol Cancer. 2021 Jan 11;20(1):14.
Motzer RJ, Tannir NM, McDermott DF, Arén Frontera O, Melichar B, Choueiri TK, Plimack ER, Barthélémy P, Porta C, George S, et al. CheckMate 214 Investigators. Nivolumab plus Ipilimumab versus Sunitinib in Advanced Renal-Cell Carcinoma. N Engl J Med. 2018 Apr 5;378(14):1277-1290.

Table 1 is available in the Supplementary Files section.

No competing interests reported.

Table1.txt

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of a novel anoikis-related gene signature and molecular subtype to predict prognosis and tumor microenvironment in clear cell renal cell carcinoma

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

METHODS

RESULTS

Discussion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1