Identification of hypoxia- and immune- based prognostic signature and validation of TFRC as a potential biomarker and therapeutic target in BLCA

doi:10.21203/rs.3.rs-3640530/v1

Download PDF

Research Article

Identification of hypoxia- and immune- based prognostic signature and validation of TFRC as a potential biomarker and therapeutic target in BLCA

https://doi.org/10.21203/rs.3.rs-3640530/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Bladder cancer is a common malignancy of the urinary system, and the survival rate and recurrence rate of patients with muscular aggressive (MIBC) bladder cancer are not ideal. Hypoxia is a pathological process in which cells acquire special characteristics to adapt to anoxic environment, which can directly affect the proliferation, invasion and immune response of bladder cancer cells. Understanding the exact effects of hypoxia and immune-related genes in BLCA is helpful for early assessment of the prognosis of BLCA. However, the prognostic model of BLCA based on hypoxia and immune-related genes has not been reported.

Purpose

Hypoxia and immune cell have important role in the prognosis of bladder cancer (BLCA). The aim of this study was to investigate whether hypoxia and immune related genes could be a novel tools to predict the overall survival and immunotherapy of BLCA patients.

Methods

Firstly, we downloaded transcriptomic data and clinical information of BLCA patients from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. A combined hypoxia and immune signature was then constructed on the basis of the training cohort via least absolute shrinkage and selection operator (LASSO) analysis and validated in test cohort. Afterwards, Kaplan-Meier curves, univariate and multivariate Cox and subgroup analysis were employed to assess the accuracy of our signature. Immune cell infiltration, checkpoint and the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm were used to investigate the immune environment and immunotherapy of BLCA patients. Furthermore, we We confirmed the role of TFRC in bladder cancer cell lines T24 and UMUC-3 through cell experiments.

Results

A combined hypoxia and immune signature containing 8 genes were successfully established. High-risk group in both training and test cohorts had significantly poorer OS than low-risk group. univariate and multivariate Cox analysis indicated our signature could be regarded as an independent prognostic factor. Different checkpoint was differently expressed between two groups, including CTLA4, HAVCR2, LAG3, PD-L1 and PDCD1. TIDE analysis indicated high-risk patients had poor response to immunotherapy and easier to have immune escape. The drug sensitivity analysis showed high-risk group patients were more potentially sensitive to many drugs. Meanwhile, TFRC could inhibit the proliferation and invasion ability of T24 and UMUC-3 cells.

Conclusion

A combined hypoxia and immune-related gene could be a novel predictive model for OS and immunotherapy estimation of BLCA patients and TFRC could be used as a potential therapeutic target in the future.

Hypoxia

immune

Immune microenvironment

Drug sensitivity

BLCA

TFRC

Bladder cancer (BLCA) is a common malignant cancer in urinary system¹. Unlike other cancer, BLCA can be divided into non-muscle-invasive (NMIBC) and muscle-invasive (MIBC) bladder cancer². Non-invasive and invasive tumors differ in clinical presentation, tumor invasiveness, pathological type and prognosis³. Besides, muscle-invasive cancer are more likely to have recurrence. As for treatment, NMIBC patients are usually treated with surgery and have bladder perfusion after surgery. For MIBC patients, radical surgery is generally selected, and adjuvant therapy such as chemotherapy and immunotherapy are decided according to postoperative pathological conditions^4,5. However, even after comprehensive treatment, the survival rate and recurrence rate of MIBC patients are still not ideal⁶. Therefore, predictors of early prognosis and immunotherapy responsiveness of bladder cancer need to be identified.

Hypoxia is a pathological process by which cells acquire specific properties to adapt to a hypoxic environment⁷. Hypoxia can affect the proliferation, invasion and immune response of bladder cancer cells⁸. This was considered to be associated with the alternative of The tumor microenvironment (TME)^9,10. Nevertheless, hypoxia can also influence the radiotherapy, chemotherapy and even the genetic instability and malignant progression of MIBC^11,12. These studies indicated that hypoxia is closely related to the prognosis of bladder cancer and thus can be thought to be a potential treatment.

TME is a complex system composed of cancer cells, extracellular matrix, immune cells and other molecules, which plays an important role in the occurrence and development of bladder cancer^13–15. It is thought the loss of immune cell function in the tumor microenvironment can free malignant tumors from immune surveillance¹⁶. In recent years, immune-related genes has attracted more and more attention from researchers. Xu et al had found immune-related genes (IRGs) was associated with prognosis in RCC¹⁷. Xiao and his colleagues had found IRGs were involved in T-cell activation, cell killing, and NK cell activity and other biological processes and were strongly related to OS¹⁸.

Up to now, the relationship between the expression of hypoxia- and immune-related genes and BLCA has not been studied in detail. In this study, a risk score model based on eight hypoxia- and immune-related genes was constructed and validated by GEO database. The AUC values and survival analysis results illustrated d the feasibility and accuracy of the risk score model. TIDE and TCIA database showed the high value of risk score model in immunotherapy response. Taking together, our results demonstrated the high value of risk score model and nomogram for the prediction of survival and immunotherapy for patients with BLCA.

Methods

The gene expression data and corresponding clinicopathological characteristics (including age, gender, survival status, survival time, tumor grade, tumor stage, TNM stage.) of bladder cancer patients were downloaded from The Cancer Genome Atlas (TCGA) database (https://gdc-portal.nci.nih.gov/), which including 412 BLCA samples and 19 normal bladder samples. GSE13507 from Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/) was downloaded as test cohort. Simple nucleotide variation (SNV) data was obtained from TCGA database, and tumor mutation burden (TMB) was defined as the total number of nonsynonymous alterations (SNVS or indels) for each patient.

Dentification of differentially Expressed HRGs and IRGs

254 hypoxia-related genes were collected from GSEA database (http://www.gsea-msigdb.org/gsea)^19,20. IMMPORT database (https://www.immport.org/home) was used to collect IRGs and 1794 IRGs were selected. “limma” R package was employed to identify the differentially expressed genes (DEGs) between tumor and normal samples with FDR < 0.05 and |logFC| > 1. These DEGs were visualized by heatmap and volcano plot. Venn diagrams were used to graphically describe the combination, intersection, and difference between DEGs, HRGs and IRGs²¹.

Pathway Enrichment Analysis

To better understand the pathways and bioactivities of these intersection genes, Kyoto Encyclopedia of Genes and Genomes (KEGG) was employed to analysis signaling pathways and Gene Ontology (GO) was used to analyze biologic processes, molecular functions, and cellular components through “clusterProfiler” R package. These analyses were aimed to determine whether the genes filtered for further investigation were indeed involved in hypoxia and immunity.

Construction of risk model based on intersection genes

First, we used univariate Cox regression analysis to identify the relationship between the expression of HRGs and IRGs and overall survival. Then the least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression were used to determine the optimal prognostic factors of the model. After that, risk score model was calculated according to the following formula:

The Coef means the coefficient of each relevant mRNA in the risk model. Based on the median risk score, TCGA cohort and GEO cohort was divided into high-risk and low-risk groups while the TCGA cohort was defined as train cohort and GEO as test cohort.

Analysis and validation of risk model

Kaplan-Meier curves were plotted to illustrate the different OS between high- and low-risk groups in train and test cohorts via “survival” and “survminer” R packages. Patients’ survival status and survival time were also estimated. Besides, progression-free survival (PFS) between subgroups of BLCA patients in test cohort was plotted via Kaplan-Meier curves. Univariate and multivariate Cox regression were conducted to assess whether risk model could be regarded as independent risk parameter. Principal component analysis (PCA) were employed to the model feasibility and accuracy. The accuracy, sensitivity, and specificity of the model were estimated by the Receiver Operating Characteristic (ROC) curves.

Identification the association of risk score model and clinicopathological features

Heatmap was plotted by “pheatmap” R package to illustrate the relationship between eight genes’ expression and different clinicopathological features including risk, age, gender, stage, grade, T, N and M. Box plots illustrated the different risk score in different clinicopathological features.

Construction of nomogram

We constructed a nomogram survival model via“rms” R package to predict the 1-, 3- and 5-year survival probability. Calibration curves, ROC curves, univariate Cox and multivariate Cox analysis were employed to illustrate the high accuracy of nomogram.

Analysis of DEGs between subgroups

The different expressed genes (DEGs) between subgroups were identified with FDR < 0.05 and |log2FC|≥1. Gene set variation analysis (GSVA) by “GSVA”R package was used to estimate the functional enrichment pathways in two subgroups.

Estimation of the immune cell infiltration, immune microenvironment, and genetic alterations analysis

The immune microenvironment including stromal score (stromal content), immune score (degree of immune cell infiltration), ESTIMATE score (a composite marker of stroma and immunity) and tumor purity was illustrated by “estimate” R package. The immune cell infiltration between two groups were calculated by CIBERSORT and visualized in box plots. The coalition of different immune cell and OS were plotted through Kaplan-Meier curves. TMB was analyzed by the “maftools” R package. The survival probability of low-TMB and high-TMB patients in subgroups was also assessed.

Checkpoint estimation and immune therapy estimation

Immune-checkpoint between high-and low-risk groups was estimated. The expression of common checkpoint including CTLA4, PD-L1, PDCD1, LAG3 and HAVCR2 between two subgroups was investigated, and their coalition with risk score was also searched. Furthermore, the tumor immune dysfunction and exclusion (TIDE) algorithm was to conducted to determine response to immunotherapy in TCGA cohort. Then we used Cancer Immune Atlas (TCIA) database (http://tcia.at/) to obtain immunotherapy files of BLCA patients and further investigated the potential predictive roles of risk score in immunotherapy response in different treatments.

Potential drugs prediction

Furthermore, we calculated the chemotherapy sensitivity of each BLCA patient according to the Genomics of Drug Sensitivity in Cancer (GDSC) database. IC50 values for each chemotherapy drug were further determined by regression analysis. IC50 values for each drug was further determined and was compared between high-risk and low-risk groups via boxplots.

Cell lines and culture conditions

T24 and UMUC3 cells are acquired from Procell Life Science & Technology Co., Lt. T24 cells are cultured with RPMI-1640 medium supplemented with 10% FBS and 1% ABAM, while UMUC3 cells are cultured in the UM-UC-3 Cells Complete Medium. All cells were maintained at 37℃ in a humidified 5% CO₂ incubator.

Edu assay

T24 and UMUC3 cells were seeded at a density of 4000cells/well into 96-well plates then transported with two independent siRNA targeting TFRC. After 48h, cells were added with EdU and continued incubating for another 2h. Then, the cells were fixed with a 4% paraformaldehyde solution for 30 min. The staining process was conducted in accordance with the manufacturer's instructions. Images were captured under microscope, and the number of positive cells was quantified utilizing imageJ software.

Wound healing assay

T24 and UMUC3 cells were seeded in 6-well plates in triplicates for each condition, two of which were treated with two independent si-TFRC. The cells were allowed to grow until full confluency was achieved. A 10 µL pipette was used to form a uniform size wound in the center of the cell monolayer for each condition, and a photograph was taken at 0 h. Cells were then observed for wound healing and photomicrograph at 48 h was taken under microscope.

Invasion assay

The cells were transfected with either control or siRNA and seeded on the Matrigel-coated PET membrane in the upper compartment. The lower compartment was filled with complete growth media and the plates were maintained at 37°C for 24 h. Cells that migrated to the lower side of membrane were fixed and stained with crystal violet. The cells were then photographed using a light microscope and the number of cells were counted using ImageJ software.

Statistical analysis

All statistical analyses were used R software (Version 4.1.3). P < 0.05 was served as the cutoff criterion.

Identification of differently expressed Hypoxia-Related Genes and Immune-Related Genes in BLCA patients

We screened 7394 differently expressed genes between normal and tumor tissue samples (Supplementary Fig. 1 and Table 1), including 2848 downregulated genes and 4546 upregulated genes with FDR value < 0.05 and |logFC| > 1 (Fig. 1A). Using public databases, we identified 254 hypoxia-related genes (HRGs) and 1793 immune-related genes (IRGs). Then we employed Venn diagram to identify intersection genes from DEGs, HRGs and IRGs. Finally, 18 genes were filtered (Fig. 1B, Supplementary Table 2) and considered the differentially expressed hypoxia- and immune-related genes. These genes were used for subsequent analysis.

Functional Analysis of Hypoxia-Related Genes and Immune-Related Genes Pathways

GO function analysis showed that these genes were involved in hypoxia, cell proliferation, steroid hormone and lipopolysaccharide (Fig. 1C). KEGG pathway analysis showed that these genes were involved in PI3K − Akt signaling pathway, MAPK signaling pathway, Rheumatoid arthritis, IL − 17 signaling pathway and TNF signaling pathway (Fig. 1D). These results indicated these genes were closely related to hypoxia and immunity.

Construction and validation of risk score model

Firstly, we used univariate Cox analysis to identify the relationship between these genes and overall survival, and 10 genes were screened (Fig. 2A). In order to optimize model, we employed LASSO regression analysis (Fig. 2B, C). Therefore, a risk score model including JUN, STC1, PROK1, TFRC, TGFB3, PLAU, PGF and SPP1 was established (Supplementary Table 3). Subsequently, we divided BLCA patients in TCGA and GEO cohorts into low and high-risk groups according to the median risk score (Fig. 2E, F). The prognosis of BLCA patients in the low-risk group was better than that in the high-risk group (Fig. 2H, I). The expression levels of the 8 hypoxia- and immune-related genes were visualized in the heatmap (Fig. 2J, K). Kaplan-Meier curves illustrated that high-risk group had worse OS than low-risk group both in train cohort and test cohort (p < 0.001 and p = 0.039 respectively) (Fig. 2D, G). High-risk group was also observed worse PFS than low-risk group in train cohort (p = 0.003) (Fig. 2L). The univariate Cox and multivariate Cox analysis were performed in the train cohort, which indicated our risk score model was an independent prognostic factor (p < 0.001) (Fig. 3A, B). PCA and t-SNE analysis illustrated our model could distinguish dimensions among different groups both in train and test cohorts (Fig. 3C, D, F, G). ROC analysis showed the AUCs for 1-, 3- and 5-year survival was 0.629, 0.668 and 0.704 respectively in train cohort and 0.681, 0.649 and 0.635 respectively in test cohort (Fig. 3E, F). These data suggested that our model could be helpful in predicting the outcome of BLCA patients.

The coalition between different clinical features and risk model

We conducted a heatmap to illustrate the coalition between different clinical features and risk model (Fig. 4A). Risk score have closely relation with age (> 60 years or ≤ 60 years), gender (female or male), grade (high or low), stage (stage I-II or stage III-IV), T (T1-2 or T3-4), N (N0 or N1-3) and M (M0 or M1) (Fig. 4B-H). These results indicated our risk score model was in good agreement with different clinical and pathological features.

Construction of predictive nomogram and calibration curves

Based on the previous results, we constructed a predictive nomogram to provide a more convenient and accurate tool that can predict the survival risk of particular BLCA patients (Fig. 4I). Our results showed that nomograms could be used as an effective tool for prognostic assessment of BLCA patients, and the calibration curves indicated our nomogram had high accuracy in BLCA patients’ survival rate (Fig. 4J).

Functional enrichment analysis of DEGs between high- and low-risk groups

With FDR value < 0.05 and |logFC| > 1, we identified 1547 differently expression genes between subgroups, including 248 downregulated genes and 1299 upregulated genes (Supplementary table 4). Then we performed GSEA, GO and KEGG analysis to show the enriched pathways. GSEA indicated the chemokine signaling pathway, cytokine-cytokine receptor interaction, ECM receptor interaction, focal adhesion and systemic lupus erythematosus were enriched in high-risk group, indicating high- risk group might eager to have metastasis (Fig. 5A, B). The results of GO and KEGG enrichment analysis also illustrated that collagen − containing extracellular matrix and PI3K-Akt pathway was activated in high-risk group (Fig. 5D, E).

Tumor mutation burden and survival analysis

The maftool package in R software is utilized to identify the top 15 mutation genes in high-risk and low-risk groups (Fig. 5C, F). Tumor mutation burden (TMB) between two groups had no significantly different (Fig. 5G). Nevertheless, subsequent analysis indicated that while combined with risk score, high TMB with low risk had better outcome and low TMB with high risk had worse outcome (Fig. 5H).

Analysis of TME, immune cell infiltration and immune-checkpoint

As for TME estimation, ESTIMATE algorithm was employed. Violin plot showed that stromal score, immune score and ESTIMATE score are significantly higher in high-risk group, indicating lower tumor purity (Fig. 6A). By CIBERSORT algorithm, we compared the difference of infiltration immune cells between high and low risk groups (Fig. 6B). The proportion of T cells CD8, T cells follicular helper, T cells regulatory (Tregs) were significantly higher in the low-risk group than in the high-risk group, while the proportion of T cells CD4 memory resting and Neutrophils were significantly higher in the high-risk group than in the low-risk group (Fig. 6C). In order to investigate their agreement with OS, survival curves were plotted. Except neutrophils, the results of coalition of immune cell with OS were in good agreement with immune cell infiltration results (Fig. 6D- G). Then we estimated the different immune functions between two groups. The heatmap indicated many immune-functions were higher enriched in high group, including T cell co − stimulation HLA and checkpoint (Fig. 6H). These results illustrated strong association between risk score model and tumor immunity. Furthermore, we identified gene expression of different checkpoint. We found that many checkpoint including CTLA4, CD274, also known as PD-L1, HAVCR4, LAG3 and PDCD1 were extremely higher expressed in high-risk group, indicating there might be different immunotherapy response between two groups (Fig. 7A). Spearman correlation analysis showed that CTLA4, HAVCR2, LAG3, PDCD1 and CD274 were significantly positively correlated with risk score (Fig. 7B- K). Generally, these genes were positively related to risk score.

Prediction of immunotherapy response and potential drugs

According to above results, we considered that our risk score model might play an important role in immunotherapy response prediction in BLCA patients. The TIDE algorithm showed that high-risk group had higher TIDE score, indicating the higher immune escape potential and poorer immunotherapy response of patients in the high-risk group (Fig. 8A- D). Therefore, we obtained immunotherapy file from TCIA database for further investigation. The violin plot demonstrated that the effect of immunotherapy was significantly improved in low-risk group when using anti-CTLA4 treatments (Fig. 8E- H). These results confirmed the strong association between risk score model and immunotherapy response in BLCA. Furthermore, we calculated the different drugs’ sensitivity of each patient based on GDSC database. Our risk score model had significant association with cisplatin, docetaxel, olaparib, staurosporine, paclitaxel, sorafenib, linsitinib and talazoparib (Fig. 8I- P). It might provide a novel prospective in BLCA treatment.

Identifying TFRC as prognosis marker for BLCA

The circus plot shows the position of the model gene in the chromosome (Fig. 9A). To identify hypoxia- and immune-related prognosis markers for BLCA, we analyzed the mutation frequency of the model genes via CNV, and the results showed that TFRC had the highest acquired mutation frequency (Fig. 9B). TFRC was significantly higher expressed in BLCA tumor samples compared with normal samples (Fig. 9C), and survival curves showed BLCA patients with low TFRC expression level had significantly longer OS compared with that with high TFRC expression level (Fig. 9D). The expression levels of TFRC in pan-carcer were also analyzed, and the box plot demonstrated differential expression of TFRC between normal and tumor tissue samples across multiple tumors (Fig. 9E). Spearman correlation analysis was conducted to demonstrate the relationship between TFRC expression and CTLA4 as well as PTCD1 (Supplementary Fig. 2A, B). This study discovered an association between TFRC and the immune microenvironment in patients with bladder cancer. This identification allows for the classification of bladder cancer patients into six immune subtypes, which aids in categorizing BLCA based on various immune response types (Supplementary Fig. 2C).

Knockdown of TFRC inhibited BLCA cell proliferation, migration and invasion

To validate the biological function of TFRC in BLCA, we knocked down TFRC using two siRNAs in T24 and UMUC3 cells. Edu staining results showed that knockdown of TFRC significantly reduced the proliferation of BLCA cells (Fig. 10A, B). The wound healing rate of T24 and UMUC3 cells was significantly inhibited by BLCA depletion, as indicated by the results of the wound-healing assay (Fig. 10C). The invasion ability of HepG2 and Hep3B cells was significantly inhibited in transwell assay following TFRC knockdown (Fig. 10D).

Bladder cancer as one of the most common malignancies worldwide which characterized as high morbidity and high mortality^22,23. Although researchers have made many attempts and explorations in the treatment of bladder cancer, the prognosis of patients with bladder cancer has not improved substantially due to high postoperative tumor recurrence rate²⁴. Therefore, novel prognostic and outcome model for BLCA needs to be identified urgently.

Hypoxia plays a vital role in cancer cell survival as it can increase tumor cell proliferation and tumor cell transition, which causing malignant phenotype transition²⁵. This might indicated that study tumor hypoxic environments could help the decision of BLCA treatment. Immunotherapy is an important for BLCA patients. Thus immune cell infiltration and immune-related genes had gained much research^26,27. Although many prognostic models have constructed and predicted survival outcomes in BLCA based on hypoxia-related genes or immune-related genes, combing hypoxia- and immune-related genes is a novel method. When these two phenotypes were nested, the prognosis of patients could be more accurately evaluated and immunotherapy could be guided.

In this study, we identified 18 differentially expressed hypoxia and immune-related genes in BLCA patients through analysis of public databases. By using Cox and LASSO regression analysis, we developed a new prognostic model containing 8 genes (JUN, STC1, PROK1, TFRC, TGFB3, PLAU, PGF and SPP1) to predict overall survival in BLCA patients. Furthermore, Kaplan-Meier curves of disease-free survival (DFS) and overall survival (OS) showed that high-risk group had worse outcome. The consistent results was observed in test cohort, indicating this hypoxia- and immune-related gene risk score model can be used as a prognostic marker for BLCA. Furthermore, we confirmed this model had strong association with virous clinical features. Besides, the impact of risk score model on the immune microenvironment, immune checkpoint, immunotherapy response, and anti-tumor drug sensitivity were also investigated.

JUN is a dimer and functions as a transcription factor and The heterodimers of JUN have a more stable and relatively stronger DNA-binding behavior^28,29. Research had confirmed the associated with the invasiveness of colorectal cancer cell³⁰. STC1 is a hypoxia-induced molecular target which promotes the progression of different types of cancer, including gastric cancer³¹, colorectal cancer³², breast cancer³³ and bladder cancer^34,35. As a ligand of the PROKR2, PROK1 can induces biological changes by transducing important molecular signals³⁶. PROK1 can influence the expression of EG-EGFR thus influence the cell proliferation and hematogenous metastasis^37–39. TFRC is considered to be an important player in intracellular iron transport, which induces the proliferation and metastasis by up-regulating the expression level of AXIN2⁴⁰. Besides, many models containing TFRC had been confirmed to have associated with prognosis of bladder cancer^41–43. In this study, we also found the role of TFRC in the proliferation of bladder cancer. We used siRNA to reduce TFRC expression in bladder cancer cells and found that TFRC knockdown inhibited tumor cell proliferation and reduced their invasive ability. AXIN2 might be one of the regulatory factors, but specific mechanism needs to be further studied.

TGFB3, also known as transforming growth factor-B3, was associated with anti-PD-L1 monoclonal antibody treatment in urothelial carcinoma⁴⁴. Inhibition of TGF-β signaling can conquer resistance to PD-1/PD-L1 blocking in cancer⁴⁵. PLAU is involved in tumor cell migration and invasion⁴⁶, and previous study had confirmed the relationship between PLAU and OS in TCGA cohort⁴⁷.PGF was also discovered to be associated with BLCA prognosis and OS^48,49. It has also been suggested that SPP1 was correlated with poor clinical outcomes and promote tumor progression by interacting with carcinogenic genes and facilitating immune cell infiltration⁵⁰. High SPP1 expression levels were strongly connected with higher stage and grade in upper tract urothelial carcinoma⁵¹.

After analyzing the different genes between subgroups, we found PI3K-Akt pathway was activated in high-risk group. PI3K-Akt pathway was a “star pathway” in bladder which regulated autophagy, apoptosis and cancer progression. According to these results, we induced that the willingness of high-risk groups to metastasize and muscle-invasive might be related to the PI3K-Akt pathway. Furthermore, we investigated immune environment between subgroups. In hypoxic environments, macrophages synthesize chemokines and cancer cells attract regulatory T cells from the circulation and suppress the antitumor responses of other T cells⁵². In our study, we also found significantly immune cell infiltration including many types of T cells.

In this study, we demonstrated improved survival outcomes in the high TMB group, which is consistent with previous studies⁵³. Groups with high TMB and lower risk scores had the best survival outcomes, whereas groups with low TMB and higher risk scores had the worst survival outcomes. Taking together, a combination of TMB and risk score could significantly improve the prediction of overall survival in bladder cancer. Other studies had found that in bladder cancer, high TMB status can increase the neoantigen burden and thus enhance the immunotherapeutic effects^54,55. Therefore, this risk score model might provide new insights into the underlying mechanism with high TMB.

Besides, tumor progression is usually affected by abnormal pathological conditions of tumor microenvironment, such as tumor-associated fibroblasts (CAFs), ECM deposition, vasodilation, and immune response suppression⁵⁶. In this study, we employed ESTIMATE to evaluate components of tumor. The increase of matrix components and immune cell components indicates that the lower the purity of the tumor, the more likely the tumor metastasis and the worse the prognosis.

Based on these results, we assume that patients in different subgroups might have different immune-checkpoint expression and response to immunotherapy. After investigation, we discovered many checkpoints including CTLA4, PD-L1, HAVCR4, LAG3 and PDCD1 had different expression, indicating different immune-response. As a novel cancer treatment option, immunotherapy gives hope to many bladder cancer patients. The correlation of immune infiltration with immunotherapy response in BLCA cases has also been reported⁵⁷. We observed patients in the high-risk group had significant lower responses than those in the low-risk group for CTLA-4-positive or both negative. TIDE results indicated high-risk group was easier to have immune escape. Furthermore, we found that low-risk group was more sensitive to cisplatin, docetaxel, olaparib, staurosporine, paclitaxel and sorafenib. In conclusion, our risk model can assess the prognosis, immune status, immunotherapy response, and drug sensitivity of BLCA cases.

However, there are certain flaws in this study. First, this study was based on public database, thus the results may be skewed for different ethnic groups and regions. Second, the particular pathways might require further investigation with in vivo and in vitro experiments.

This 8-genes risk score model was an accurate and reliable tool for predicting clinical outcomes, immunotherapy response, and anti-tumor drug sensitivity of BLCA patients, providing novel prospective for BLCA treatment.

Author Contribution

Tang.RH. and Wang.HR. wrote the main manuscript text.Liu.JY. and Song.LQ. prepared figures 9-10. Hou.HM., LIU.M., Wang.JY. and Wang.JL provided guidance on the writing of the manuscript.All authors reviewed the manuscript.

Lenis, A. T., Lec, P. M., Chamie, K. & Mshs, M. D. Bladder Cancer: A Review. JAMA 324, 1980–1991 (2020).
Cumberbatch, M. G. K. et al. Epidemiology of Bladder Cancer: A Systematic Review and Contemporary Update of Risk Factors in 2018. Eur. Urol. 74, 784–795 (2018).
Yousef, P. G. & Gabril, M. Y. An update on the molecular pathology of urinary bladder tumors. Pathol. Res. Pract. 214, 1–6 (2018).
Soukup, V. et al. Prognostic Performance and Reproducibility of the 1973 and 2004/2016 World Health Organization Grading Classification Systems in Non-muscle-invasive Bladder Cancer: A European Association of Urology Non-muscle Invasive Bladder Cancer Guidelines Panel Systematic Review. Eur. Urol. 72, 801–813 (2017).
Taylor, J., Becher, E. & Steinberg, G. D. Update on the guideline of guidelines: non-muscle-invasive bladder cancer. BJU Int. 125, 197–205 (2020).
Jo, Y. Y., Yea, J. W., Park, J., Oh, S. A. & Park, J. W. Optimized Adaptive Radiotherapy with Individualized Plan Library for Muscle-Invasive Bladder Cancer Using Internal Target Volume Generation. Cancers 14, 4674 (2022).
Choudhry, H. & Harris, A. L. Advances in Hypoxia-Inducible Factor Biology. Cell Metab. 27, 281–298 (2018).
Jing, X. et al. Role of hypoxia in cancer therapy by regulating the tumor microenvironment. Mol. Cancer 18, 157 (2019).
LaGory, E. L. & Giaccia, A. J. The ever-expanding role of HIF in tumour and stromal biology. Nat. Cell Biol. 18, 356–365 (2016).
McAllister, S. S. & Weinberg, R. A. The tumour-induced systemic environment as a critical regulator of cancer progression and metastasis. Nat. Cell Biol. 16, 717–727 (2014).
Lodhi, T., Song, Y. P., West, C., Hoskin, P. & Choudhury, A. Hypoxia and its Modification in Bladder Cancer: Current and Future Perspectives. Clin. Oncol. R. Coll. Radiol. G. B. 33, 376–390 (2021).
Mao, X., Nanzhang, null, Xiao, J., Wu, H. & Ding, K. Hypoxia-Induced Autophagy Enhances Cisplatin Resistance in Human Bladder Cancer Cells by Targeting Hypoxia-Inducible Factor-1α. J. Immunol. Res. 2021, 8887437 (2021).
da Costa, J. B. et al. Molecular tumor heterogeneity in muscle invasive bladder cancer: Biomarkers, subtypes, and implications for therapy. Urol. Oncol. 40, 287–294 (2022).
Hu, J. et al. 5mC regulator-mediated molecular subtypes depict the hallmarks of the tumor microenvironment and guide precision medicine in bladder cancer. BMC Med. 19, 289 (2021).
Warrick, J. I. et al. Intratumoral Heterogeneity of Bladder Cancer by Molecular Subtypes and Histologic Variants. Eur. Urol. 75, 18–22 (2019).
Lawson, K. A. et al. Functional genomic landscape of cancer-intrinsic evasion of killing by T cells. Nature 586, 120–126 (2020).
Xu, W.-H. et al. Prognostic value and immune infiltration of novel signatures in clear cell renal cell carcinoma microenvironment. Aging 11, 6999–7020 (2019).
Xiao, Y. et al. Characterization of the immune related lncRNAs in bladder cancer to aid immunotherapy. Front. Immunol. 13, 941189 (2022).
Sc, W. et al. Relation of a hypoxia metagene derived from head and neck cancer to prognosis of multiple cancers. Cancer Res. 67, (2007).
Subramanian, A. et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U. S. A. 102, 15545–15550 (2005).
Jia, A., Xu, L. & Wang, Y. Venn diagrams in bioinformatics. Brief. Bioinform. 22, bbab108 (2021).
Antoni, S. et al. Bladder Cancer Incidence and Mortality: A Global Overview and Recent Trends. Eur. Urol. 71, 96–108 (2017).
Lenis, A. T., Lec, P. M., Chamie, K. & Mshs, M. D. Bladder Cancer: A Review. JAMA 324, 1980–1991 (2020).
Mossanen, M. The Epidemiology of Bladder Cancer. Hematol. Oncol. Clin. North Am. 35, 445–455 (2021).
Sun, X. et al. Identification and validation of a hypoxia-related prognostic and immune microenvironment signature in bladder cancer. Cancer Cell Int. 21, 251 (2021).
Zhang, J. et al. Identification of Immune-Related Subtypes and Construction of a Novel Prognostic Model for Bladder Urothelial Cancer. Biomolecules 12, 1670 (2022).
Feng, Z.-H. et al. m6A-immune-related lncRNA prognostic signature for predicting immune landscape and prognosis of bladder cancer. J. Transl. Med. 20, 492 (2022).
Lei, S. et al. JUND/linc00976 promotes cholangiocarcinoma progression and metastasis, inhibits ferroptosis by regulating the miR-3202/GPX4 axis. Cell Death Dis. 13, 967 (2022).
Su, C.-C. et al. Involvement of AMPKα and MAPK-ERK/-JNK Signals in Docetaxel-Induced Human Tongue Squamous Cell Carcinoma Cell Apoptosis. Int. J. Mol. Sci. 23, 13857 (2022).
Martinez-Ledesma, E., Verhaak, R. G. W. & Treviño, V. Identification of a multi-cancer gene expression biomarker for cancer clinical outcomes using a network-based algorithm. Sci. Rep. 5, 11966 (2015).
Wang, Y. et al. Stanniocalcin–1 promotes cell proliferation, chemoresistance and metastasis in hypoxic gastric cancer cells via Bcl–2. Oncol. Rep. 41, 1998–2008 (2019).
Rezapour, S. et al. STC1 and NF-κB p65 (Rel A) is Constitutively Activated in Colorectal Cancer. Clin. Lab. 62, 463–469 (2016).
Chang, A. C.-M. et al. STC1 expression is associated with tumor growth and metastasis in breast cancer. Clin. Exp. Metastasis 32, 15–27 (2015).
Fang, Z., Tian, Z., Luo, K., Song, H. & Yi, J. Clinical significance of stanniocalcin expression in tissue and serum of gastric cancer patients. Chin. J. Cancer Res. Chung-Kuo Yen Cheng Yen Chiu 26, 602–610 (2014).
Sun, J. et al. STC1 is a Novel Biomarker Associated with Immune Characteristics and Prognosis of Bladder Cancer. Int. J. Gen. Med. 14, 5505–5516 (2021).
Zhang, Y. et al. Risk Factors for Synchronous Peritoneal Metastases in Colorectal Cancer: A Systematic Review and Meta-Analysis. Front. Oncol. 12, 885504 (2022).
Yan, X. et al. EG-VEGF silencing inhibits cell proliferation and promotes cell apoptosis in pancreatic carcinoma via PI3K/AKT/mTOR signaling pathway. Biomed. Pharmacother. Biomedecine Pharmacother. 109, 762–769 (2019).
Nakazawa, T., Goi, T., Hirono, Y. & Yamaguchi, A. Prokineticin 1 protein expression is a useful new prognostic factor for human sporadic colorectal cancer. Ann. Surg. Oncol. 22, 1496–1503 (2015).
Goi, T. et al. Expression of prokineticin-receptor2(PK-R2) is a new prognostic factor in human colorectal cancer. Oncotarget 6, 31758–31766 (2015).
Huang, Y. et al. TFRC promotes epithelial ovarian cancer cell proliferation and metastasis via up-regulation of AXIN2 expression. Am. J. Cancer Res. 10, 131–147 (2020).
Sun, J. et al. Identification of a Novel Ferroptosis-Related Gene Prognostic Signature in Bladder Cancer. Front. Oncol. 11, 730716 (2021).
Liang, Y. et al. A novel survival model based on a Ferroptosis-related gene signature for predicting overall survival in bladder cancer. BMC Cancer 21, 943 (2021).
Luan, J.-C. et al. A novel signature constructed by ferroptosis-associated genes (FAGs) for the prediction of prognosis in bladder urothelial carcinoma (BLCA) and associated with immune infiltration. Cancer Cell Int. 21, 414 (2021).
Peng, M. Immune landscape of distinct subtypes in urothelial carcinoma based on immune gene profile. Front. Immunol. 13, 970885 (2022).
de Streel, G. et al. Selective inhibition of TGF-β1 produced by GARP-expressing Tregs overcomes resistance to PD-1/PD-L1 blockade in cancer. Nat. Commun. 11, 4545 (2020).
Chen, Q. et al. Screening and identification of hub genes in pancreatic cancer by integrated bioinformatics analysis. J. Cell. Biochem. 120, 19496–19508 (2019).
Jin, Y. et al. An Integrated Analysis of Prognostic Signature and Immune Microenvironment in Tongue Squamous Cell Carcinoma. Front. Oncol. 12, 891716 (2022).
Qu, G. et al. Development of a prognostic index and screening of prognosis related genes based on an immunogenomic landscape analysis of bladder cancer. Aging 13, 12099–12112 (2021).
Na, L. et al. Identification of 9-Core Immune-Related Genes in Bladder Urothelial Carcinoma Prognosis. Front. Oncol. 10, 1142 (2020).
Tu, Y., Chen, C. & Fan, G. Association between the expression of secreted phosphoprotein - related genes and prognosis of human cancer. BMC Cancer 19, 1230 (2019).
Li, Y. et al. Identification of plasma secreted phosphoprotein 1 as a novel biomarker for upper tract urothelial carcinomas. Biomed. Pharmacother. Biomedecine Pharmacother. 113, 108744 (2019).
Petrova, V., Annicchiarico-Petruzzelli, M., Melino, G. & Amelio, I. The hypoxic tumour microenvironment. Oncogenesis 7, 10 (2018).
Patel, V. G., Oh, W. K. & Galsky, M. D. Treatment of muscle-invasive and advanced bladder cancer in 2020. CA. Cancer J. Clin. 70, 404–423 (2020).
Jiang, T. et al. Tumor neoantigens: from basic research to clinical applications. J. Hematol. Oncol.J Hematol Oncol 12, 93 (2019).
Bellmunt, J., Powles, T. & Vogelzang, N. J. A review on the evolution of PD-1/PD-L1 immunotherapy for bladder cancer: The future is now. Cancer Treat. Rev. 54, 58–67 (2017).
Wang, J. et al. Identification and verification of an immune-related lncRNA signature for predicting the prognosis of patients with bladder cancer. Int. Immunopharmacol. 90, 107146 (2021).
Yang, Z. et al. Immune escape mechanisms and immunotherapy of urothelial bladder cancer. J. Clin. Transl. Res. 7, 485–500 (2021).

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of hypoxia- and immune- based prognostic signature and validation of TFRC as a potential biomarker and therapeutic target in BLCA

Status:

Version 1

Abstract

Figures

Introduction

Data collection

Dentification of differentially Expressed HRGs and IRGs

Pathway Enrichment Analysis

Construction of risk model based on intersection genes

Analysis and validation of risk model

Identification the association of risk score model and clinicopathological features

Construction of nomogram

Analysis of DEGs between subgroups

Estimation of the immune cell infiltration, immune microenvironment, and genetic alterations analysis

Checkpoint estimation and immune therapy estimation

Potential drugs prediction

Cell lines and culture conditions

Edu assay

Wound healing assay

Invasion assay

Statistical analysis

Results

Identification of differently expressed Hypoxia-Related Genes and Immune-Related Genes in BLCA patients

Functional Analysis of Hypoxia-Related Genes and Immune-Related Genes Pathways

Construction and validation of risk score model

The coalition between different clinical features and risk model

Construction of predictive nomogram and calibration curves

Functional enrichment analysis of DEGs between high- and low-risk groups

Tumor mutation burden and survival analysis

Analysis of TME, immune cell infiltration and immune-checkpoint

Prediction of immunotherapy response and potential drugs

Identifying TFRC as prognosis marker for BLCA

Knockdown of TFRC inhibited BLCA cell proliferation, migration and invasion

Discussion

Conclusion

Declarations

Author Contribution

References

Additional Declarations

Supplementary Files

Status:

Version 1