Analysis of low expression of ferroptosis-related gene DECR1 on poor prognosis and immune cell defects of bladder urothelial carcinoma

Lan-Yu Wang ,Song-Yun Zhao ,Chun-Yang Chen ,Jian-Feng Shao*

1Department of Urology,Wuxi People's Hospital Affiliated to Nanjing Medical University,Wuxi 214000,China.2Department of Neurosurgery,Wuxi People's Hospital Affiliated to Nanjing Medical University,Wuxi 214200,China.

Abstract Background: Ferroptosis is an iron dependent form of cell death,which plays an important role in the pathogenesis of a variety of urinary malignancies.The down-regulation of DECR1 gene causes the accumulation of polyunsaturated fatty acids (PUFAs),increases the susceptibility to lipid peroxidation,and finally leads to cell death.Methods: We first searched the data to find reductase 1(DECR1)expression in bladder uroepithelial carcinoma and healthy surrounding tissues in the Cancer Genome Atlas(TCGA),and we then confirmed DECR1 expression with additional independent cohorts in the Gene Expression Omnibus(GEO)database and the Human Protein Atlas(HPA).In order to determine the link between DECR1 expression and clinical traits and overall survival (OS),as well as to create nomograms,multivariate analysis and Kaplan Meier survival curves were utilized.Through the use of an online string website,the network of proteins that interact with DECR1 was created.Through an online string website,the protein network interacting with DECR1 was built.Finally,we looked at the association between aggressive immune cells and the marker genes that belong to them and DECR1 expression.Results: When compared to normal tissues,bladder tumor tissues had higher DECR1 expression (P=0.002).Low DECR1 expression was linked with tumor grade and stage in tumor cells.BLCA patients with low DECR1 expression had a lower overall survival than BLCA patients with high DECR1 expression,according to a survival study (P=0.008).The protein network’s HSD17B4 and DECR1 connections are crucial.The expression of regulatory T cells,regulatory B cells,and their markers were decreased when DECR1 was missing in bladder cancer.Conclusion: Decreased DECR1 expression was associated with BLCA progression,poor prognosis and impaired infiltration of some immune cells.

Keywords: ferroptosis;reductase 1;bladder urothelial carcinoma;immune cell infiltration;bioinformatics analysis

Introduction

Worldwide,the incidence of bladder cancer is the fourth and ninth most common malignant tumor of the urinary system in men and women,respectively [1].Every year,between 350,000 and 380,000 new cases are reported globally.The two clinical phenotypes of bladder cancer are muscle-invasive and non-muscle-invasive[2],with over 70% of bladder cancers diagnosed as non-muscle invasive[3].

TURBt surgery and radical cystectomy are the mainstays of treatment for patients with bladder cancer[4].As a result,cancer cells have been discovered to accumulate a lot of iron,and tumors are more likely to induce ferroptosis iron.There are several pathways involved in ferroptosis,an iron-dependent kind of cell death brought on by the buildup of lipid peroxides,including mevalonate,iron,sugar,and lipid metabolism [5,6].Ferroptosis has been associated with cancer,blood disorders,immune system disorders,and heart disease [7].In order to promote growth,Cancer cells need more iron ions than normal cancer cells.Therefore ferroptosis has been considered promising for cancer treatment [8].New clinical approaches are urgently required to improve the sensitivity to radiation and chemotherapy in advanced bladder uroepithelial carcinoma.There is increasing evidence in recent years that ferroptosis plays a significant role in a number of cancers,such as lung cancer,colon cancer,and glioblastoma [9-11].Excess intracellular iron is stored in ferritin,which may lead to iron overload when iron storage in ferritin is reduced,or cellular iron uptake is increased.Excess iron generates excess reactive oxygen species (ROS) through the Fenton reaction,which leads to ferroptosis[12].

Coenzyme 2,4-Diene,it has been demonstrated that a DECR1,which is a rate-limiting enzyme for the oxidation of PUFAs,plays a variety of functions in many forms of cancer,including upregulated expression in prostate cancer and down-regulated expression in breast cancer[13-15].In prostate cancer,downregulation of the DECR1 gene led to PUFA buildup,increased sensitivity to lipid peroxidation,and induced ferroptosis [16].Previous studies have reported that lower free iron concentrations significantly favor the proliferation of bladder cancer cells,and thus inhibition of ferroptosis may be closely associated with the proliferation of bladder cancer cells.Increased ferroptosis may also promote the effect of immunotherapy,and the presence of positive feedback between ferroptosis and immunotherapy synergistically kills cancer cell death [17,18].

In this work,we used databases including TCGA,GEO,and Human Protein Atlas to download and evaluate the link between DECR1 expression and clinical data of BLCA patients and overall survival.We then looked at the connection between DECR1 expression and invading immune cells using the TIMER database and the cybersport method.Additionally,the STRING website was used to examine the protein interaction network of DECR1,and the results revealed that:low DECR1 expression is a sign of poor prognosis and is connected to immune cell infiltration in BLCA.Because of this,DECR1 loss may lessen ferroptosis in tumor cells and thereby lessen the immunological impact.Combining immunotherapy with the targeting of DECR1 may be a therapeutic approach to cause ferroptosis in BLCA patients.

Materials and methods

Raw data acquisition

The UCSC Xena website (https://xenabrowser.net/datapages/)provided the DECR1 expression data and TCGA pan-cancer data.TCGA (https://cancergenome.nih.gov/) provided clinical and gene expression data for BLCA patients.All samples with no expression data or insufficient clinical data were eliminated.414 BLCA patients’clinical information and RNA-Seq gene expression FPKM (fragments per million kilobases) were kept and later examined (Table 1).Since the database is open to the public,the local ethics committee’s permission is not necessary.

Table 1 Relationship between DECR1 expression and clinical traits of patients

One of the largest gene chip collections in the world,the National Center for Biotechnology Information’s (NCBI)(https://www.ncbi.nlm.nih.gov/geo/)GEO database is a comprehensive gene expression repository.Additionally,findings of protein immunohistochemistry for both tumor tissues and normal human tissues are accessible on this website.

Protein-protein interaction (PPI) analysis

The online tool (STRING) website (https://string-db.org/) is a resource for finding proteins that interact with one another,and it contains a significant amount of integrated information about these interactions.We collected the PPI network data after importing DECR1 into the web program STRING,and we then chose the interaction networks with confidence values larger than 0.7.

Immune cell infiltration analysis

We used the R package “CIBERSORT” to analyze the relative expression levels of 22 immune cell types to determine the extent of immune cell infiltration in bladder cancer [19].The CIBERSORT inverse fold product technique uses vector regression and the LM22 dataset to estimate cell type proportions in tumor samples with a variety of cell types.22 immune cell abundance matrices for each sample were obtained by analyzing RNA-Seq(FPKM format)data from COAD samples.The CIBERSORT output withP ＜0.05 indicates significance.

In the meanwhile,10,897 samples from the TCGA database are available on the public website TIMER (http://cistrome.org/TIMER/),which is intended to determine the degree of tumor cell immune infiltration.Through the TIMER database,we evaluated the relationship between DECR1 gene expression and tumor purity as well as the relationship between DECR1 expression and the content of six different types of infiltrating immune cells in BLCA patients,including CD8+T cells,CD4+T cells,B cells,dendritic cells,macrophages,and neutrophils.

Genetic correlation analysis

GEPIA(https://gepia.cancer-pku.cn/index.html) is an online database with tumors and data.In the GEPIA database,we investigated the relationship between DECR1 expression and multiple markers of immune cells in BLCA.In addition,we used TIMER data to validate genes significantly associated with DECR1 expression in GEPIA.

Statistical analysis

R (v.3.6.3) software was used to perform statistical analysis.As the data showed normal distribution between BLCA groups,differences in DECR1 expression levels between normal tissues and tumors were tested using the Wilcoxon rank sum test.The Kolmogorov-Smirnov test was used to analyze the relationship between DECR1 expression and clinicopathological features.In this study,BLCA patients were divided into a high DECR1 expression group and a low DECR1 expression group according to the median expression of the DECR1 gene.To investigate whether DECR1 expression levels affected the clinical outcome of BLCA patients,we compared survival differences using Kaplan-Meier survival curves.

To further determine the impact of DECR1 expression in BLCA patients,we used univariate and multifactorial Cox regression analyses to calculate the relationship between DECR1 expression levels and other clinical traits and patient OS.

Clinical characteristics of the high and low DECR1 expression groups were screened for prognostic variables using univariate and multivariate Cox regression(R package“survival”)and then studied to identify independent prognostic factors for model construction.The nomogram model was created using the R package “rms”.The ROC curves for the column line graphs were plotted using the R package“survival ROC”.

Results

Expression analysis of DECR1 in BLCA

In order to investigate the expression levels of DECR1 mRNA in various tumor types,we included RNA sequencing data and thoroughclinical prognostic information resources from 414 BLCA cases and 19 normal tissue samples from the TCGA database in our analysis.Figure 1 in the TCGA database shows that DECR1 gene expression levels were considerably higher in tumor samples compared to healthy BLCAC tissues (P=0.002),and these findings were confirmed in the GEO database (P=0.035,P=0.051) (Figure 2).In the human protein atlas,the expression of the DECR1 protein was higher in BLCA tumor tissues than in healthy tissues.Analysis of the correlation between DECR1 expression and clinicopathological parameters in BLCA patients showed no significant correlation between DECR1 mRNA levels and age (P=0.984),N stage,and M stage.And there were significant differences in DECR1 expression levels in T2 and T3 stages(P＜0.001),histological malignancy (P＜0.001),tumor type (P＜0.001),pathological stage (P＜0.001),and gender (P=0.031).It was evident that DECR1 expression was lower in stage T3,high-grade,stage III-IV,and non-papillary BLCA.

Independent prognostic analysis of DECR1 in BLCA

Overall survival and disease-specific survival (DSS) were also shorter in the BLCA tumor group with low DECR1 expression according to Kaplan-Meier survival analysis (Figure 3A,P=0.008 Figure 3B,P=0.006).In the univariate Cox model,NCOA4 expression,pathological grade,tumor type,and age were all factors associated with OS treatment outcomes in BLCA patients (Figure 3C,Pall ＜0.05).In the multifactorial regression analysis,NCOA4 expression,pathological grade,and age were independent factors associated with OS (Figure 3D,Pall ＜0.05).

Age,tumor stage,gender,and DECR1 expression were all included in the nomogram column line plot(Figure 3E).The prognostic model’s projected OS and actual overall survival were compared using calibration plots,and the findings demonstrated the accuracy of the column line graph’s prediction results (Figure 3F).

Construction of PPI network

We used the STRING web tool to determine the PPI network of DECR1 proteins to discover the interactions of these proteins in BLCA progression.The top 10 proteins and corresponding gene names and co-expression scores are shown in Figure 4 and Table 2.These genes include HADHB,ACAA2,HSD17B4,CALB1,ACADM,PTTG1,PRKCB,PRKCA,ECI1,and OSGIN2.Interestingly,HADHB can modulate the action of ferroptosis inhibitors in lymphoma,and in the OncomineTM database,mRNA expression of PRKCB was higher in tumor tissues than in normal tissues [20,21].

Correlation analysis of DECR1 expression and infiltrating immune cells

Next,we looked into the relationship between DECR1 expression in BLCA and tumor immune infiltration.By using Spearman’s correlation analysis and the R package “cybersport,” we discovered a correlation between DECR1 expression levels and immune cell content in the COAD tumor microenvironment.Additionally,we visualized the relationship between the expression of 22 immune cells and NCOA4 using a lollipop plot(Figure 5A).The 22 immune cells in the high and low DECR1 expression groups in the violin plot were similarly distinct(Figure 5B).Correlation score and difference analysis findings revealed that M0 macrophages were more abundant in the low DECR1 expression group and exhibited a negative correlation.Still,regulatory T cells and B cells were more abundant in the high DECR1 expression group and showed a positive correlation.

Using the TIME database,we also examined the relationship between DECR1 expression and tumor purity with regard to six different invading immune cell types (CD8+T cells,CD4+T cells,B cells,dendritic cells,macrophages,and neutrophils) (Figure 5 C).The findings revealed a significant positive connection (r=0.239,P=4.00e-06) between the amounts of invading B cells and DECR1 levels.In BLCA,DECR1 expression likewise showed a correlation (r=0.317P=4.47e-10) with tumor purity.

Figure 1 Expression status of DECR1 in uroepithelial carcinoma of the bladder.(A)Expression levels of DECR1 in different cancer tissues and corresponding normal tissues.(B) High expression level of DECR1 in BLCA compared to normal tissues.(C-E) No statistical difference between DECR1 expression levels and age,pathological N stage,and pathological M stage.(F-J) Lower DECR1 expression was associated with higher pathological T-stage (P ＜0.001),poorer histological malignancy (P ＜0.001),poorer tumor type (P ＜0.001),higher pathological stage (P ＜0.001),and female gender (P=0.031).DECR1,reductase 1.

Figure 2 (A-B) Expression analysis of DECR1 gene in GEO datasets GSE13507 and GSE52519.DECR1 mRNA expression was higher in BLCA than normal tissues.(C-D)The level of NCOA4 protein in human protein profile was higher in urinary bladder epithelial carcinoma than in normal tissue(Antibody HPA023160,10X).DECR1,reductase 1.

Figure 3 Prognostic value of DECR1 in BLCA.(A-B) KaPlan-Meier survival curves for overall survival (OS) and disease-specific survival (DSS) in BLCA patients with high and low DECR1 expression.(C-D) Univariate and multifactorial regression analysis of DECR1 and other clinicopathological symptoms and OS in BLCA patients.(E) Nomogram for predicting 1-year,3-year,and 5-year survival probability (OS) in BLCA patients.(F) Calibration curves for predicting 1-year,3-year,and 5-year survival probability (OS) in BLCA patients.OS,overall survival;DSS,disease-specific survival;DECR1,reductase 1.

Figure 4 DECR1-interacting proteins in BLCA.DECR1,reductase 1.

Figure 5 (A) Results based on cibersort algorithm,association of DECR1 expression with tumor infiltrating immune cells,violin plot comparing the percentage of immune cells in high and low expression groups.(B)Lollipop plot of the correlation between 22 immune cells and NCOA4 expression based on the cibersort algorithm.(C) Correlation of DECR1 expression with infiltrating immune cells in BLCA based on TIMER database.DECR1,reductase 1.

Additionally,we investigated the association between DECR1 and the expression of a number of immunological markers,including CD8+T cells,T cells,B cells,M1/M2 macrophages,tumor-associated macrophages (TAM),neutrophils,monocytes,N.K.cells,and D.C.cells,using the GEPIA and TIMER databases (Table 3).The findings demonstrated that most immune cells,including T cells,B cells,TAMs,M1/M2 macrophages,and monocytes,expressed DECR1 in BLCA.

Table 2 Annotation of DECR1-interacting proteins and their co-expression scores

Table 3 Correlation analysis of DECR1 with immune cell markers in TIMER and GEPIA

Table 3 Correlation analysis of DECR1 with immune cell markers in TIMER and GEPIA (Continued)

Conclusion

Prostate cancer cells with deficient 2,4-dienoyl coenzyme a DECR1,a mitochondrial enzyme involved in the breakdown of PUFAs,exhibit slowed tumor development[22].DECR1 was significantly upregulated in prostate tumor samples and in a mouse model of denudation resistance but was downregulated in breast cancer.This suggests that DECR1 plays a diverse role in different types of cancer [23].At the same time,DECR1 deficient prostate cancer cells accumulate more unsaturated fatty acids and lead to a strong endoplasmic reticulum stress response.Thus DECR1 has a great potential role in the control of redox homeostasis [24].In bladder cancer,no relevant mechanisms have been studied.In a previous study on DECR1,high DECR1 expression was associated with shorter disease-free survival in prostate cancer patients [25].We also observed higher DECR1 expression in urinary bladder epithelial carcinoma,but low DECR1 expression was associated with a higher tumor grade and a worse prognosis(Figure 3).These findings suggest that DECR1 expression inbladder cancer may influence tumorigenesis and the progression of ferroptosis,but its deletion may affect bladder cancer patients’prognoses through various mechanisms.

Therefore,we investigated the role of DECR1 expression in the development of BLCA and its prognosis based on TCGA,GEO,and Human Protein Atlas databases.According to STING software analysis,Hydroxyalkyl coenzyme A dehydrogenase beta (HADHB) is also a critical protein interacting with DECR1 molecules in BLCA.Also,HADHB can modulate the action of ferroptosis inhibitors in lymphoma.It is a major challenge to realize the use of tumor biomarkers to help diagnose bladder cancer and determine the choice of treatment options.

Our research revealed a strong correlation between the invasion of B cells,tumor-associated macrophages,and CD4+/CD8+T cells in BLCA and the degree of DECR1 expression.TAM markers such as CCL2,CD80,CD86,and CCR5 were linked to DECR1 expression(Table 3).TAMs are significant elements of the tumor microenvironment that regulate the growth of the tumor and are frequently linked to a poor prognosis in solid tumors [26].This raises the possibility that DECR1 regulates TAM polarization and affects bladder cancer patients’ prognoses.The production of IL-10 by tumor-associated regulatory B cells may adversely influence the prognosis of bladder cancer,which is interesting because we discovered a substantial association between DEDCR1 and B-cell biomarkers [27].These results further suggest that DECR1 is a key molecule linking the ferroptosis process and immunotherapy in BLCA.Along with the development of cancer therapy,as time goes on,ferroptosis is viewed as a potential and successful combination treatment,and DECR1 expression may be a potential new marker for both ferroptosis and immunotherapy [28].

The absence of information from several tumors and normal sample sets,the heterogeneity of the data,and platform variations are some of the study’s limitations.Our investigation demonstrated that the only independent predictive variables for OS in BLCA patients were BLCA stage and patient age with DECR1 expression.Furthermore,since the TCGA database only contains data on white and black people,extrapolating these findings to other races is challenging.For the advancement of this field,additional research is required on the validation of bioinformatic prediction results,including Western blot or immunohistochemical staining assays for proteins and functional analysis of DECR1 in promoting iron cell apoptosis and immunotherapy in vivo and in vitro.Currently,there is very little experimental and clinical research on the link between bladder cancer and ferroptosis.Only a few bioinformatic analyses pose the possibility that genes related to ferroptosis serve as bladder cancer treatment targets and molecular indicators [25,29].More carefully planned research is required to investigate the possible molecular processes and clinical consequences of ferritin-mediated treatment for urologic malignancies in light of the aforementioned data.