Six‐transmembrane epithelial antigen of prostate 3 (STEAP3) is a potential prognostic biomarker in clear cell renal cell carcinoma that correlates with M2 macrophage infiltration and epithelial–mesenchymal

Abstract Background The six‐transmembrane epithelial antigen of the prostate 3 (STEAP3) is a metalloreductase, which is essential for iron uptake. Existing literature has shown that STEAP3 may perform an important role in the onset and progression of tumors. Nonetheless, a complete pan‐cancer investigation of the prognostic significance and immune properties of STEAP3 is currently unavailable. Aims As part of our investigation into the possible functions of STEAP3 in malignancies, we conducted a comprehensive analysis to examine the prognostic value and immune features of STEAP3 in human pan‐cancer. Methods and Results R and Cytoscape programs were applied to analyze and visualize the data. The expression of STEAP3 in both cell lines and tissues was measured utilizing a variety of approaches. Using the shRNA knockdown technique, we tested the viability of the A498 and 786‐O cell lines and validated their functions. Both CCK‐8 and transwell assays were conducted to estimate cell proliferation and invasion. The expression of STEAP3 was substantially elevated and was shown to be linked to prognosis in the majority of malignancies, notably in clear cell renal cell carcinoma (ccRCC). In addition, the expression of STEAP3 was shown to have a strong correlation with immune infiltrates, which in turn triggered the recruitment and polarization of M2 macrophages in ccRCC. The protein STEAP3 shows promise as a predictor of responses to immune‐checkpoint blockade (ICB) therapy. Positive links between STEAP3 and the epithelial‐mesenchymal transition (EMT), the p53 pathway, and the immune‐associated pathways were also found in the enrichment analysis. Our results illustrated that the STEAP3 expression level was substantially elevated in ccRCC tissues and suggested that it could stimulate EMT in ccRCC by downregulating CDH1. Conclusion In a diverse range of cancers, STEAP3 might serve as a biomarker for determining the prognosis as well as a predictor of immunotherapy responsiveness. STEAP3 is a novel biological marker for determining prognosis, and it also performs a remarkable function in the promotion of tumor growth in ccRCC by enhancing invasion and EMT, as well as by triggering the recruitment and polarization of M2 macrophages.


Conclusion:
In a diverse range of cancers, STEAP3 might serve as a biomarker for determining the prognosis as well as a predictor of immunotherapy responsiveness.
STEAP3 is a novel biological marker for determining prognosis, and it also performs a remarkable function in the promotion of tumor growth in ccRCC by enhancing invasion and EMT, as well as by triggering the recruitment and polarization of M2 macrophages.

| INTRODUCTION
Cancer has a huge impact on human health worldwide. Cancer contributes to a large proportion of all deaths and ranks second as the leading contributor to mortality in the world after angiocardiopathy. 1 Cancer therapy has rapidly progressed in the past decade. Immunotherapy, particularly immune-checkpoint blockade (ICB), has emerged as a promising cancer treatment approach. 2 However, a large population of patients fail to respond to immunotherapy or relapse, because of the heterogeneity of immune cell infiltrates. In light of this, the discovery of immune-related biological markers that are both specific and sensitive continues to be an important step toward the development of innovative diagnostic and therapeutic techniques.
Iron homeostasis is crucial for maintaining cellular functions. Iron metabolism is closely related to cancer development 3 and subtle changes can influence the local microenvironment and function of immune cells. 4 Several researchers have reported the role of the sixtransmembrane epithelial antigen of the prostate (STEAP) family, a type of iron metabolism-associated protein in different cancers. 5,6 STEAP family member 3 (STEAP3) performs a pivotal function in iron uptake as a metalloreductase by reducing ferric iron to ferrous iron. 7,8 STEAP3 may serve a remarkable role in tumorigenesis. It can promote hepatocellular carcinoma cell proliferation by enhancing RAC1-ERK-STAT3 signaling. 9 Increased levels of STEAP3 expression are associated with a worse prognosis in a diverse variety of tumor types, including glioblastoma, breast carcinomas, and renal cell carcinoma. [10][11][12] Further, STEAP3 may elicit specific cytotoxic lymphocyte responses and can be a promising candidate for prostate cancer immunotherapy. 13 Nevertheless, a systematic review of STEAP3's predictive significance and immune features across all cancer types has not been documented.
Given STEAP3's specific functions in malignancies and its therapeutic potential, the prospects of its use as a pan-cancer biomarker are particularly significant. Therefore, we conducted a systematic analysis of the prognostic value and immune characteristics of STEAP3. Furthermore, we performed pathway analysis to explore its   Table S1.
We also obtained 946 cell lines expression matrix of tumors from the Cancer Cell Line Encyclopedia (CCLE; https://portals. broadinstitute.org/ccle) data set. 15 The GSE15641 (Platform: GPL96) microarray data were obtained from the Gene Expression Omnibus (GEO) data set (http:// www.ncbi.nlm.nih.gov/geo). 16 We analyzed a total of 33 cases with ccRCC and 23 control kidney tissue samples. The unprocessed data were obtained in the format of MINiML files, and after extraction, they were subjected to log2 transformation to achieve normalization. The data obtained from the microarrays were normalized with the help of the function "normalize quantiles" found in the "prepro-cessCore" R package. Using the annotation information of the normalized data in the platform, the probes were transformed into gene symbols. Multiple-gene-matching probes were excluded from these sets of data. The average expression value of the gene measured by multiple probes was calculated as the final expression value.

| Protein-level analysis
The Human Protein Atlas (HPA) database (https://www.proteinatlas. org/) is one of the most frequently visited human proteome databases. 17 We collected the STEAP3 expression data by analyzing the immunohistochemical (IHC) staining images obtained from the HPA database, and 190 samples were included for further analysis.

| Differential expression analysis
The variances in STEAP3 expression levels between tumors and adjoining tissues were compared using R software (v3.4.1). A bar graph depicting STEAP3 expression was generated using the "ggplot2" R package. For a comparison between two groups, we utilized the Wilcoxon test, whereas, for a comparison among multiple groups, the Kruskal-Wallis test was applied.

| Survival prognosis analysis
Using the TCGA data set, we assessed STEAP3's prognostic significance for a subset of cancers and prognosis types via univariate Cox regression analysis by applying the "forestplot" R package, we visually represented the p-value, hazard ratio (HR), and 95% confidence interval (CI) for each variable in a forest plot.
Then, we applied Kaplan-Meier (KM) analysis using "survival" and "survminer" R packages to show the variation in overall survival (OS) between patients with high-and low-expression groups to assess the prognostic significance of STEAP3. Finally, survival rates were compared among the groups using the Logrank test.

| TISDB database
TISDB (http://cis.hku.hk/TISIDB/) is a web-based platform that combines data from multiple sources. We used this database to determine whether STEAP3 expression is associated with pathological grades or clinical stages of ccRCC. We entered "STEAP3" in the "Quick Search" module, and links between STEAP3 and clinical features were obtained from the "Clinical" module. 18 2.6 | cBioPortal database cBioPortal (http://www.cbioportal.org) is a usable and reliable portal to perform cancer genomics analysis. The mutational profile of STEAP3 in various tumors was analyzed using this database. "TCGA, PanCancer Atlas" was chosen to serve as the cohort. Thereafter, we entered "STEAP3" in the "Quick Search" module. The "cancer type summary" and the "mutation" modules both presented STEAP3 alteration sites, types, and numbers. 19

| Association of STEAP3 expression with tumor mutational burden and Microsatellite Instability
Tumor mutational burden (TMB) is defined as the count of somatic nonsynonymous mutations per megabase. 20 Microsatellites are repeats of a few nucleotides (1-6 bp) that are distributed across the genome. 21 We used TCGA-derived RNA-seq data to determine TMB and microsatellite instability (MSI) scores and then analyzed the relationships between TMB, MSI, and STEAP3 expression utilizing Spearman's correlation analysis.

| Immune cell infiltration and immune checkpoint analyses
The reliability of the immune score assessment results was evaluated using the "immuneeconv" R package, which incorporates two latest algorithms: TIMER and CIBERSORT. Using TCGA-derived data, we

| Tumor immune syngeneic MOuse database analysis
Tumor immune syngeneic MOuse (TISMO) database is an immunotherapy-response predictor database based on the syngeneic mouse tumor model. We used this database to access whether STEAP3 expression is related to immunotherapy response. We entered "STEAP3" in the "Gene" module and obtained results from the "Gene in Vivo" module, 22 and 832 ICB studies were included in further analysis.

| Correlation between STEAP3 and drug response
Gene set cancer analysis (GSCA) (http://bioinfo.life.hust.edu.cn/GSCA/ #/) is an easy-to-use web server for the visualization of gene and drug sensitivity correlation. We used the GSCA database to examine the link between STEAP3 expression and sensitivity to drugs. We entered "STEAP3" in the "Drug" module. "GDSC drug sensitivity and expression correlation" and "CTRP drug sensitivity and expression correlation" were chosen for the correlation analysis. 23 Additionally, the IC50 data for 732 small-molecule drugs were included in further analysis.

| Enrichment analysis of genes associated with STEAP3
STRING (https://cn.string-db.org/) is a network mapping tool based on the STRING protein interaction algorithm. Applying this database, we mapped out STEAP3's potential network of interaction, which was visualized using the Cytoscape software. We entered "STEAP3" in the "Protein by name" module. The main parameters that were established were as follows: network edge meaning ("evidence"), minimum threshold score for interaction ["low confidence (0.150)"], and a maximum number of interactors to show ("no more than 50 interactors" in the first shell). 24 Enrichment analysis was conducted on the top 50 genes that were chosen based on the potential interaction network of STEAP3.
These genes were searched in the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontology (GO) database using the Metascape tool (http://metascape.org). We submitted the list of the top 50 genes on the home page and obtained the results in the "Analysis Report Page." 25 By using TCGA-derived data, we used single sample gene set enrichment analysis (ssGSEA) to explore the pathways affected by STEAP3 with the "GSVA" R package. First, we obtained the Hallmark signatures from the Molecular Signatures Database. 26 Afterward, we computed an enrichment score for each sample and conducted a Spearman correlation analysis of these scores and STEAP3 expression. Finally, the "ggplot2" R package presented the results in the form of a bubble plot, which included a summary of all the findings.

| Single-cell level pathway exploration
CancerSEA (biocc.hrbmu.edu.cn/CancerSEA/home.jsp) is a strong functional analysis software based on data obtained by sequencing single cells. In total, 93 475 cancer single cells from 74 single-cell data sets were included for further analysis. We used this database to explore whether STEAP3 expression is related to 14 cancer-associated functional states. We entered "STEAP3" in the "Search" module and obtained results from the "Relevance" module. 27

| Candidate ceRNA regulatory network prediction
The ceRNA regulatory network of STEAP3 was predicted using PITA, RNA22, miRmap, microT, miRanda, PicTar, TargetScan, and StarBase gene prediction programs. 28 The relationship between miRNAs and lncRNAs was visualized in the Cytoscape tool.

| RNA extraction and cDNA synthesis
To obtain total RNA from cells and tissues, we followed the directions provided by the manufacturer of the SPARKeasy bacterial/cell RNA kit (AC0202, Sparkjade). Thereafter, 2 the T series Multi-Block thermal cycler (LongGene) was then utilized to synthesize first-strand cDNA from 2 μg of total RNA.

| Real-time quantitative PCR
The STEAP3 relative mRNA expression was detected using Real-time quantitative PCR (RT-qPCR), with the fast start universal SYBR green master kits (S2008, BR Healthcare) and CFX96TM real-time PCR System (Bio-Rad). GADPH represented the endogenous normalization reference. All the primers are listed in Table 2. The primary anti-STEAP3 (PA5-62202, Thermo Scientific) and

| Cell culture and antibody
anti-alpha-tubulin (A11126, Thermo Scientific) rabbit antibodies were diluted following the manufacturer's specifications.

| Western blotting
The RIPA buffer was utilized for the total protein extraction. The proteins were separated using SDS-PAGE (10% resolving gel) before transferring them onto PVDF membranes. The next step involved incubating the membranes for an entire night with diluted primary antibodies followed by 30 min of incubation with the secondary antibody. The gel imaging system (Tianneng) was used to detect the signal strength.

| Immunohistochemical analysis
Paraffin blocks of tumor tissues were used to prepare unstained slides for immunohistochemistry. In brief, paraffin sections were immunostained using a streptavidin peroxidase procedure after heat repair of the antigen. The signal was detected using a 3,3 0 -diaminobenzidine solution and hematoxylin and the slides were observed using a microscope (Leica, Germany).

| shRNA and lentivirus transduction
Specific shRNA against STEAP3 (GCUUCAUGCCCGUGGACAUTT) and its corresponding control were purchased from JiKai Gene.
shSTEAP3 was transduced into 786-O and A498 cells, and puromycin was used to select cells showing stable transfection for 2 weeks. The detailed procedures were mentioned in the operation manual provided by JiKai Gene.

| Cell proliferation assay
In 96-well plates (Corning), cell seeding was conducted at a density of 3000 cells per well. and cultured for 0, 24, 72, and 120 h. Thereafter, each well was supplied with 10 μL CCK-8 reagent (Beyotime) and subjected to incubation for 2 h at 37 C. The microplate reader (Bioteck) was utilized to measure the absorbance at 450 nm.

| Cell migration assay
We added 20 000 cells with 200 μL serum-free medium (SFM) in the upper chamber and 600 μL complete medium in the lower chamber of 24-well plates (Corning). After 16 h, the cells were stained with 1% crystal violet for 5 min after being fixed for 20 min with 4% tissue fixative.
Lastly, the migrated cells were observed using a microscope (Leica).

| Statistical analysis
Both SPSS 23 and R software v4.03 were utilized in the analyses of statistical data. 3 | RESULTS

| Pan-cancer analysis of STEAP3 expression
To gain more insights into STEAP3's expression profile, we evaluated its mRNA expression in 946 tumor cell lines from the CCLE database.   Figure 1A). Then, we measured the protein expression of STEAP3 in various cancers in the HPA database. IHC data of 190 samples revealed that STEAP3 is highly expressed in lung, renal, urothelial, prostate, breast, and liver cancers ( Figure 1B). We subsequently made a comparison in STEAP3 expression between tumors and normal samples to probe how STEAP3 exerts its  Figure 1C). The additional details are included in Table S3.

| Prognostic values of STEAP3
We conducted a survival analysis for STEAP3 in pan-cancer. In patients with UVM, GBM, KIRC, LGG, MESO, KIRP, OV, SARC, and ACC, STEAP3 was a risk factor for unfavorable OS (Figure 2A).  Figure 2G,H). Notably, our results showed that STEAP3 was expressed at a higher level in ccRCC individuals with lymph node and distance metastases ( Figure 2I,J).

| The characteristics of STEAP3 mutations
Considering the abnormal STEAP3 expression observed in different cancers, we analyzed mutations in the STEAP3 genes. We found that followed by mutation ( Figure 3A). cBioPortal database analysis showed that missense mutations were the main type of STEAP3 gene mutations in pan-cancer ( Figure 3B). TMB and MSI are the potential predictors of the response to ICB therapy. 29  OV ( Figure 3C). STEAP3 expression and MSI were positively correlated in STAD, BRCA, and LUSC and negatively correlated in READ and PRAD ( Figure 3D).  Figure 4D).

| STEAP3 and the response to cancer therapy
We further analyzed the predictive role of STEAP3 in ICB therapy.
We identified the expression of STEAP3 after immunotherapy using the TISMO database. 832 ICB studies were included for further analysis. The results showed that the STEAP3 expression levels in different models were significantly upregulated in the responders but not in the non-responders after ICB therapy (especially anti-CTLA4 antibody treatment), which implied that the upregulated STEAP3 was related to a better response for ICB therapy ( Figure 5A). The additional details are included in Table S4.
Therefore, the potential prediction ability of STEAP3 was further confirmed in the ICB response.
Further, we analyzed the relationships between STEAP3 expression and 732 common chemotherapy agents and small molecular compounds. Combining mRNA expression and drug sensitivity data from GDSC and CTRP, correlation analysis was conducted to find sensitive drugs ( Figure 5B). We did not find any significantly high correlation (r > .5) between STEAP3 expression and drug IC50 in most cases.

F I G U R E 5 (Continued)
GSK-J4 and tozasertib were the only two drugs having a significantly high correlation with STEAP3 expression. The additional details are shown included in Table S5.

| Molecular function analysis of STEAP3 in pan-cancer
The 50 most correlated genes were obtained from the STRING database to further analyze the functional characteristics of STEAP3, and the potential interaction network was constructed ( Figure 6A).
GO/KEGG enrichment analysis showed that most related terms correlated with iron metabolism ( Figure 6B). Next, we analyzed functional characteristics based on the hallmark terms (based on the ssGSEA score as the standard). We discovered a positive link between STEAP3 and immune-associated pathways in diverse cancers, particularly in PRAD, ACC, LGG, KIRC, LIHC, LUSC, GBM, and THYM. We also found that p53 pathways and EMT were positively related to STEAP3 ( Figure 6C). Then, utilizing single-cell sequencing data from the CancerSEA database, we investigated the correlation between STEAP3 expression and 14 cancer functional states, and 93 475 cancer single cells from 74 single-cell data sets were included for further analysis. STEAP3 was positively related to hypoxia, EMT, invasion, metastasis, and stemness in most tumors ( Figure 6D). For ccRCC, a strong positive link between STEAP3 expression and hypoxia and stemness was observed, and a negative correlation with DNA repair and inflammation was observed. We explored the upstream ncRNA regulatory network of STEAP3 to investigate the unknown role of STEAP3. Forty-five candidate miRNAs and 56 lncRNAs were predicted using various prediction programs ( Figure S2A,C). Considering expression, survival, and correlation analyses, the most potential ceRNA networks may be SNHG4/hsa-miR-204-5p/STEAP3 axis and LINC00997 or CYTOR/hsa-miR-27b-3p/STEAP3 axis in ccRCC ( Figure S2B,D).

| STEAP3 promotes invasion and EMT of ccRCC
STEAP3 mRNA levels were elevated in the ccRCC tumor tissues relative to the normal tissues according to the TCGA and GSE15641 databases ( Figure 7A,B). These results were confirmed via RT-qPCR, western blotting, and immunohistochemistry. We discovered that STEAP3 was expressed at a high level in patients with ccRCC F I G U R E 6 (Continued) ( Figure 7C-E). Next, we confirmed our results through in vitro experiments. We used specific STEAP3-targeting shRNAs to knock down STEAP3 in the 786-O and A498 cells. STEAP3 knockdown was confirmed using western blotting. The results of CCK-8 and transwell assays demonstrated that the knockdown of STEAP3 inhibited the proliferation and invasion of the 786-O and A498 cells ( Figure 7G,H).
Thereafter, changes in the expression of EMT marker genes (CDH1, CDH2, and VIM) in 786-O and A498 cells before and after STEAP3 knockdown were detected using RT-qPCR. We found that STEAP3 knockdown significantly lowered CDH1 levels and elevated CDH2/VIM levels in 786-O and A498 cells ( Figure 7I).

| DISCUSSION
Dysregulation of iron metabolism is closely related to cancer develop-

ment. Recent studies on new proteins involved in iron metabolism
have shed light on the interaction between iron and cancer, [30][31][32] and these studies also suggest that proteins involved in iron metabolism may be multifunctional and contribute to the development of tumors through non-iron metabolic pathways. 33 Iron is an important factor in EMT, metastasis, and immunomodulation in human cancers. 4,34,35 STEAP3, an iron reductase, was associated with the infiltration of immune cells in hepatocellular carcinoma and induced PD-L2 expression. 36 It may induce the formation of a tumor immune microenvironment in ccRCC. 37 STEAP3 may play a role in cancer development, progression, and immunotherapy; however, few studies have explored its significance in cancer research.
First, we assessed STEAP3 expression in pan-cancer cells. Data indicated that STEAP3 is relatively highly expressed in lung, renal, urothelial, prostate, breast, and liver cancers based on the HPA database. Compared with normal tissues, STEAP3 is significantly upregulated in 22 cancer types in TCGA database, suggesting that STEAP3 may function as a key regulator in cancers. Next, we examined the prognostic value of STEAP3. Analyses of patients' survival status confirmed that STEAP3 was a risk factor in the onset and progression of many different types of cancer. In ccRCC, STEAP3 was expressed at high levels in tumor tissues suggesting poor prognosis and correlated significantly with both clinical stages and pathological grades. Moreover, we discovered that STEAP3 was expressed at high level in ccRCC with lymph nodes and distant metastasis. Taken together, STEAP3 may be involved in tumor metastasis in ccRCC. DNA mutation is one of the critical factors during oncogenesis. 38 However, we found that low frequency of alteration in the STEAP3 gene. This may because the highly conserved nature of pathways of iron metabolism.
Some researchers reported that STEAP3 are naturally processed CTL epitopes possessing anti-prostate cancer reactivity in vivo and exhibited great potential for immunotherapy. 13 In most malignancies, Extensive research has demonstrated the significant involvement ncRNAs in tumorigenesis by competing for endogenous RNA. [43][44][45] We predicted SNHG4/hsa-miR-204-5p /STEAP3 axis and LINC00997 or CYTOR/hsa-miR-27b-3p/STEAP3 axis as the most potential ceRNA networks in ccRCC. The role of these miRNAs and lncRNAs in other cancers also suggests their possible role in ccRCC.
Hsa-miR-204-5p and hsa-miR-27b-3p perform an anticancer role in other cancer types. Laryngeal squamous cell carcinoma invasion and metastasis may be inhibited by Hsa-miR-204-5p. 46 Hsa-miR-27b-3p can inhibit invasion, migration and EMT in gastric cancer 47 and suppress glioma development via targeting YAP1. 48 SNHG4 sponges miR-204-5p and promotes RCC cell proliferation and invasion. 49 The lncRNA CYTOR performs a remarkable function in promoting the development of digestive system tumors. 50 Nonetheless, there are several drawbacks to our investigation.
First, the clinical sample size was relatively small and may not be sufficient to fully describe the problem. Second, the immune infiltration-related analysis was predicted by algorithms based on RNA-seq data, and we could not obtain the real immune cell infiltration data. Although we verified the results obtained from bioinformatic analysis using in vitro experiments as much as possible, more data are required. However, despite these limitations, we still obtained partially reliable conclusions, and we will further validate the available results and explore possible mechanisms in subsequent studies.

| CONCLUSIONS
Herein, we conducted an in-depth pan-cancer investigation of STEAP3's role in cancer prognosis and immunology, demonstrating that STEAP3 could potentially serve as a predictive biological marker as well as a predictor of the responsiveness to immuno-

CONFLICT OF INTEREST STATEMENT
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

DATA AVAILABILITY STATEMENT
The unique contributions made in the work are described in the article as well as the Supplementary Material. Any further questions can be directed to the corresponding authors.

ETHICS STATEMENT
The

SUPPORTING INFORMATION
Additional supporting information can be found online in the Supporting Information section at the end of this article.