Skip to main content
NCBI home page
The World Journal of Men's Health logoLink to The World Journal of Men's Health
. 2025 Apr 23;44(1):136–149. doi: 10.5534/wjmh.240298

Single-Cell Analysis Combined with Mendelian Randomization Identifies Genes Associated with Prostate Cancer Cells

Weizhuo Wang 1,*, Kaiyu Lu 1,*, Xi Zhang 2, Kai Wei 3, Gang Cheng 1, Zheng Tang 1, Mingwei Zhan 4, Caibin Fan 1,✉, Xu Fu 1,✉
PMCID: PMC12798317  PMID: 40583027

Abstract

Purpose

Prostate cancer is the second most common male cancer, with incidence increasing with age. When prostate cancer extends beyond the prostatic capsule, treatment options and prognosis change significantly. This study aims to investigate prognostic genes related to capsular invasion in prostate cancer by integrating single-cell data with Mendelian randomization (MR) analysis.

Materials and Methods

Single-cell sequencing data from six prostate cancer cases were obtained from the Gene Expression Omnibus (GEO) database. Cell clustering and annotation were performed using R, and high-dimensional weighted gene co-expression network analysis (hdWGCNA) identified differentially expressed genes in advanced-stage cancer. Single nucleotide polymorphism loci corresponding to these genes were retrieved from the UK Biobank (UKB), and MR exposure data were acquired from the ukb-b-13348 dataset. MR analysis assessed the impact of hdWGCNA-identified genes. Clinical and gene expression data from TCGA and GEO were analyzed using univariate Cox regression to evaluate gene effects on prognosis. Cellular functional experiments and immunohistochemistry assessed gene expression and function in prostate cancer.

Results

We employed the Seurat package for quality control and integration of single-cell data from four patients. hdWGCNA identified three modules, from which 200 genes were selected. The combined analysis of eQTL and MR revealed that TMEM59, JUNB, NECTIN2, OSBPL10, ATF3, and WLS may have relevant associations with prostate cancer. Further investigation using TCGA and GEO data suggested that TMEM59 might act as a protective factor in prostate cancer. Cellular experiments confirmed that TMEM59 knockdown enhanced the proliferation and invasion capabilities of prostate cancer cells. Immunohistochemistry demonstrated a significant decrease in TMEM59 expression in both normal and tumor tissues, particularly in the tumor group.

Conclusions

These findings suggest that TMEM59 may play a crucial role in the progression of prostate cancer and could serve as a prognostic predictor and therapeutic target for the disease.

Keywords: Biomarkers, Mendelian randomization analysis, Prostatic neoplasms, Single-cell gene expression analysis, Tumor microenvironment

INTRODUCTION

Prostate cancer is a malignant tumor originating from glandular epithelial cells. It ranks as the second most common cancer in men worldwide, with higher incidence rates in developed countries [1]. The incidence is also age-related, with prostate cancer being the most common cancer in men aged 70 years and above [2]. Clinically, prostate cancer is staged based on tumor size and invasion of the prostatic capsule (T staging). Generally, T1 and T2 prostate cancers are confined to the prostate without invasion of the prostatic capsule or surrounding tissues (seminal vesicles or bladder). Once invasion of the capsule and surrounding tissues occurs, it is categorized as T3 or T4 stage [3]. The transition in clinical staging often leads to changes in the treatment approach for patients.

For localized prostate cancer (T1, T2), the treatment is usually more conservative, and for lower-stage patients (such as T1, T2a), active surveillance may even be considered, avoiding surgical intervention [4]. However, once locally advanced prostate cancer is diagnosed, treatment options may include radical prostatectomy, radiotherapy, and androgen deprivation therapy (ADT), depending on the patient’s expected lifespan. The prognosis and quality of life for patients significantly deteriorate in the case of invasive prostate cancer [5]. Therefore, determining whether there is invasion of the prostatic capsule, i.e., the boundary between T2 and T3 stages, holds great significance in clinical practice.

In the process of tumor invasion and metastasis, the tumor microenvironment plays a crucial role. Single-cell analysis provides researchers with a precise cellular perspective to investigate the intercellular interactions of tumors and identify specific cellular subgroups [6]. Notably, high-dimensional weighted gene co-expression network analysis (hdWGCNA) in single-cell sequencing enables us to perform weighted gene co-expression network analysis at the single-cell level [7]. This method allows the construction of co-expression networks across multiple cellular and spatial scales, facilitating the identification of highly correlated gene modules that may include disease-related gene sets. Moreover, owing to the continuous expansion of various omics information, especially the pioneering development of genomics, a wealth of data offers researchers additional references and research resources. Particularly noteworthy in recent years are genome-wide association studies (GWAS) data, providing an opportunity to observe diseases from the perspective of single nucleotide polymorphisms (SNPs) [8]. Mendelian randomization (MR) is a statistical method that uses genetic variants as instrumental variables to infer causal relationships between risk factors and disease outcomes. By leveraging the random assortment of alleles at conception, MR minimizes confounding and reverse causation, providing robust evidence for potential causal effects. Notably, GWAS data can be linked to patients’ exposures, gene expression, and protein expression, among other influencing factors. Serving as intermediate tools, investigating the relationships between these influencing factors and diseases, genes can be associated with disease occurrence and progression through their expression quantitative trait loci (eQTL) and MR [9].

The purpose of this study is to utilize existing single-cell data, categorize prostate cancer cases based on the presence or absence of capsule invasion, and subsequently employ hdWCGNA for co-expression module identification. This approach aims to identify expression gene modules between different groups, and these genes will be linked to their eQTL information. Ultimately, this study aims to identify key genes associated with the prognosis of prostate cancer. Validation will be conducted through transcriptomics, cell experiments, and immunohistochemistry (IHC).

MATERIALS AND METHODS

1. Data collection

Download single-cell data from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/), MR data, and corresponding eQTL information from the UK Biobank database (https://gwas.mrcieu.ac.uk/datasets/). Simultaneously, obtain bulk sequencing data for prostate cancer and corresponding clinical data from TCGA (https://portal.gdc.cancer.gov/) and GEO databases. Organize the data into transcripts per million (TPM) format and standardize the datasets.

2. Processing and annotation of single-cell data

Our analysis primarily relies on the R programming language (4.2.2). The basic analysis for single-cell data is conducted using the Seurat package (4.3.0), with the FindIntegrationAnchors fuction utilized for the integration of single-cell data to remove batch effects. We employed the NormalizeData function for data normalization and the FindVariableFeatures function to identify highly variable genes. After that, we used ScaleData fuction for data centering, and then applied RunPCA, FindNeighbors, FindClusters, and RunUMAP for dimensionality reduction. Additionally, the single R package (2.0.0) is employed for annotating cell types. The quality control criteria for single-cell data include: nFeature_RNA > 200 & nFeature_RNA < 5000 & percent.mt < 10 [10].

3. Extraction of epithelial cells and hdWGCNA

In prostate cancer, the primary source is epithelial cells. In single-cell data, epithelial cells are considered as tumor cells. The subset function is employed for extraction, and simultaneously, modular genes of single-cell data are obtained using hdWGCNA package (0.1.1.9010). Temporal analysis is planned based on the monocle package (2.24.0) [11], while cell communication is investigated using the cellchat package (1.6.1) [12].

4. hdWGCNA genes Mendelian randomization analysis

After selecting modules using hdWGCNA, the top 50 genes from each corresponding module are extracted. The eQTL information for these genes is then downloaded from the UK Biobank (UKB) database. Subsequently, outcome variables are obtained from ukb-b-13348 dataset, and the twosampleMR package (0.5.10) is used for MR analysis of eQTL and exposure outcomes. Transcriptomic data and corresponding clinical data for prostate cancer patients are downloaded and organized from TCGA and GEO databases. Univariate Cox regression analysis is employed for the selection of survival-related genes, with a significance threshold set at 0.05. Reverse Mendelian data are sourced from the UKB.

5. Real-time reverse transcription PCR

Following the instructions of the PrimeScript™ RT reagent Kit (Perfect Real Time, cat. no. RR047A, 200 reactions; Takara), reverse transcription was conducted using SYBR Premix Ex TaqII (catalog no. RR820A, Takara, 200 Rxns) for qRT-PCR. Several sequences were utilized: TMEM59 forward primer ′5-GGCAGAACTCATCAGGTCGCT-3′ and reverse primer ′5-GCATTCTTGGCATCAGGGACA-3′; β-actin forward primer ′5-AGCCATGTACGTAGCC ATCCA-3′, and reverse primer ′5-TCTCCGGAGTCCATCACAATG-3′.

6. Cell counting kit-8 experiment

According to the standard procedure, a cell counting kit-8 (CCK8) assay was conducted to assess cell proliferation. LNCaP and PC3 cells (cultured at a density of 2×105) were cultured in an incubator (Labselect). Subsequently, they were seeded in a 96-well plate with 2×104 cells per well, 100 µL of cell suspension, and five replicates for each cell line. To each well, 10 µL of CCK8 solution was added. After adding the reagent, the culture plate was gently shaken to assist mixing, and then incubated for 72 hours. The absorbance values at 450 nm were automatically measured at various time points using a microplate reader.

7. Cell proliferation assay and transwell assay

Following standard procedures, a colony formation assay was conducted to assess the proliferation capacity of cells. LNCaP and PC3 cells, cultured at a density of 2×105, were placed in the upper chamber (Labselect). LNCaP cells, originally derived from a human lymph node metastatic lesion of prostatic adenocarcinoma, express the androgen receptor and retain many characteristics of differentiated prostate epithelial cells. PC3 cells, established from a bone metastasis of a grade IV prostatic adenocarcinoma, are characterized by high metastatic potential, particularly to the bone, and are commonly used to study the mechanisms underlying metastasis and the progression of prostate cancer post-ADT. Both cell lines are extensively utilized in prostate cancer research to understand different aspects of prostate cancer biology [13]. For invasion experiments, the upper chamber was pre-coated with Matrigel (BD Biosciences). Different culture media were employed: one without fetal bovine serum for the upper chamber and another with 10% fetal bovine serum for the lower chamber. After 12 hours, non-migratory cells were removed, and the invasive prostate cancer cells were fixed, stained, and counted using an inverted microscope.

8. Immunohistochemistry staining

According to the standard protocol, IHC was performed on pathological sections from 11 prostate cancer patients, including 7 tumor tissues and 4 normal tissues. In brief, tissues were fixed in 4% paraformaldehyde solution or 10% neutral-buffered formalin, fixed overnight at room temperature, and subsequently embedded in paraffin following standard procedures. Sections were deparaffinized in xylene, rehydrated in decreasing concentrations of ethanol (100%, 90%, 80%, 75%) for 3 minutes each, and antigen retrieval was achieved by microwave heating in sodium citrate buffer. Subsequently, sections were blocked with 5% bovine serum albumin and incubated with primary antibodies: anti-LDHA (#3582#3, diluted 1:100; Cell Signaling Technology), anti-cd33 (#17425, diluted 1:100; ProteinTech Group), anti-lox-1 (#11837, diluted 1:200; ProteinTech Group), or anti-c-Rel (#sc-6955, diluted 1:100; Santa Cruz Biotechnology). Goat anti-rabbit secondary antibodies labeled with horseradish peroxidase were used for detection. The antibody binding was visualized using the DAB substrate kit (Invitrogen). Images were acquired using an inverted microscope (Olympus IX71). The H-score was calculated using the formula: H-score=Σ (PI×I)=(percentage of weakly stained cells×1)+(percentage of moderately stained cells×2)+(percentage of strongly stained cells×3), where the H-score was recorded as a continuous variable.

9. Ethics statement

All experiments in this study were conducted in compliance with relevant guidelines and regulations. Ethical approval for the study was obtained from the Ethics Committee of the Second Affiliated Hospital of Soochow University, under approval number JDLK202205901. All participants were fully informed about the study and provided their consent, with written informed consent obtained from each individual. The cells used in this study were obtained from commercial cell lines; therefore, ethical approval is not applicable.

RESULTS

1. Data collection

We obtained single-cell data from tumor tissues of four prostate cancer patients from the GSE193337 dataset (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE193337). This dataset includes clinical samples from these four patients, comprising two patients with stage T2 prostate cancer and two patients with stage T3 prostate cancer. The clinical samples from these four patients were analyzed using single-cell sequencing technology, totaling 18,195 cells. The data was grouped based on clinical information. SNP information was downloaded from the IEU GWAS database (https://gwas.mrcieu.ac.uk/datasets/ukb-b-13348/), with the dataset named ukb-b-13348, which includes SNP information for 462,933 patients, of whom 3,269 are prostate cancer patients. Transcriptomic data was downloaded from TCGA-PRAD (https://portal.gdc.cancer.gov/) and GSE116918 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE116918), organized into TPM format, and standardized. TCGA-PRAD contains RNA sequencing data and clinical information for 432 prostate cancer patients, while GSE116918 includes RNA sequencing data and clinical information for 249 patients.

2. Data preprocessing and annotation of single-cell data

We conducted quality control and preprocessing of the GSE193337 dataset using the Seurat package. Ultimately, we filtered and obtained 6,780 cells. After completing the basic analysis of the single-cell data, we identified a total of 20 cell clusters (Fig. 1A, 1B). Further annotation of each cell subtype was performed using the single R package with the HumanPrimaryCellAtlasData as the reference gene set, leading to the classification of seven cell populations: endothelial cells, epithelial cells, monocytes, NK cells (natural killer cells), T cells, B cells, and SM cells (smooth muscle cells) (Fig. 1C, 1D). Subsequent cell grouping based on staging was conducted, considering that the prostate primarily originates from epithelial cells, where epithelial cells were regarded as tumor cells. Finally, we utilized the subset function to extract all epithelial cells, totaling 1,856 epithelial cells.

Fig. 1. Single-cell data analysis. (A) UMAP dimensionality reduction of single-cell data, displayed by cell clusters. (B) UMAP dimensionality reduction of single-cell data, displayed by cell types. (C) UMAP dimensionality reduction of single-cell data, displayed by cell clusters, and grouped according to clinical information. (D) UMAP dimensionality reduction of single-cell data, displayed by cell types, and grouped according to clinical information. UMAP: uniform manifold approximation and projection, NK cell: natural killer cell, SM cells: smooth muscle cell.

Fig. 1

Open in a new tab

3. Extraction of epithelial cells and hdWGCNA

We extracted epithelial cells and grouped them based on the presence or absence of capsular invasion, specifically into T2 stage (non) and T3 stage (invasion) (Fig. 2A, 2B). Simultaneously, we performed dimensionality reduction clustering on these epithelial cells. We observed that the proportions of four cell subtypes (1, 2, 3, 5 clusters) were higher in the invasion group (Fig. 2C). To investigate whether the cell communication of these subtypes differs from others, we further grouped epithelial cells into IS_Epi (1, 2, 3, 5 clusters) and Other_Epi (other clusters). Subsequently, we randomly selected 2000 cells for cell communication analysis and found significant differences in communication levels between cell subtypes (Fig. 2D). Notably, IS_Epi, the subtype with a higher expression proportion in the invasion group, exhibited more active cell communication levels (Fig. 2E, 2F). We observed that pathways such as MK (Midkine), EGF (epidermal growth factor), TWEAK (tumor necrosis factor-related weak inducer of apoptosis), Wnt, and GDF (growth differentiation factor) were significantly enhanced in the IS_Epi group (Fig. 2G, 2H). Consequently, we performed hdWGCNA analysis to identify co-expressed gene modules specific to the IS_Epi group, resulting in the identification of 13 modules (Fig. 3A–3C). We then selected three modules—turquoise, brown, and tan—comprising a total of 150 genes (Fig. 3D). Afterwards, we conducted pseudotemporal analysis on these 150 genes, observing their changes along the developmental and differentiation trajectory of the cells (Fig. 3E).

Fig. 2. Epithelial cell subgroup selection. (A) Epithelial cell cluster grouping under UMAP dimensionality reduction. (B) UMAP dimensionality reduction of epithelial cell clustering displayed by clinical group. (C) Cell group proportions, presented by cell clusters. (D) Overall cell communication of the epithelial subgroup with increased total in invasive prostate cancer and the remaining subgroups. (E) Bubble plot of overall cell communication of the epithelial subgroup with increased numbers in invasive prostate cancer and the remaining subgroups. (F) Number of cell communications of the epithelial subgroup with increased numbers in invasive prostate cancer and the remaining subgroups. (G, H) Cell communication status of each cell type. IS Epi: epithelial cells from clusters 1, 2, 3, and 5, Other Epi: epithelial cells from all other clusters except 1, 2, 3, and 5, UMAP: uniform manifold approximation and projection, NK cell: natural killer cell.

Fig. 2

Open in a new tab

Fig. 3. High-dimensional weighted gene co-expression network analysis (hdWGCNA). (A) hdWGCNA SoftPower Selection for choosing the most appropriate SoftPower value. (B) hdWGCNA Dendrogram showing the enrichment of each module. (C) Uniform manifold approximation and projection (UMAP) dimension reduction showing the expression of each module’s genes in cell subpopulations. (D) Expression of module genes across clusters. (E) The pseudo-time analysis for gene identification.

Fig. 3

Open in a new tab

4. Mendelian randomization analysis of hdWGCNA genes

After extracting the relevant genes, we obtained their eQTL information from the GWAS database. We identified the exposure outcome data (ukb-b-13348) to observe the impact of genes on the exposure outcome. After conducting MR analysis, we found causal relationships between genes such as ATF3, JUNB, NECTIN2, OSBPL10, TMEM59, and WLS with the exposure outcome (Fig. 4A). Due to the small size of the identified OR for genes, we suspected it might be attributed to the fewer cases in the outcome compared to the control samples in ukb-b-13348. Specifically, there were 3,269 cases of tumors and 459,664 cases of controls in ukb-b-13348. For further causal relationship inference, we use RNA sequencing data followed by univariate Cox regression analysis. We downloaded TCGA-PRAD and GSE116918 datasets to determine the relationship between gene expression and prognosis. Due to the limited number of cases with death as the clinical endpoint in TCGA and GSE116918 databases, and considering the good prognosis of prostate cancer, we selected biochemical recurrence as the clinical prognosis endpoint. After extracting the gene expression information from patients, we performed a univariate Cox regression (Fig. 4B, 4C). We integrated univariate Cox regression results from the TCGA and GSE116918 databases and found that only TMEM59 consistently had a p-value less than 0.05 and showed a consistent effect on biochemical recurrence across both databases. Subsequently, we conducted a reverse MR study on TMEM59 using ukb-b-13348 data and found no causal relationship between the occurrence of prostate cancer and TMEM59 (Fig. 4D).

Fig. 4. Identification of key gene. (A) Mendelian randomization of genes identified by hdWGCNA (high-dimensional weighted gene co-expression network analysis), (B, C) Univariate Cox regression of Mendelian randomization results for TCGA-PRAD and GSE116918 datasets. (D) Reverse Mendelian analysis of the TMEM59. nsnp: number of single nucleotide polymorphisms, OR: odds ratio, CI: confidence interval.

Fig. 4

Open in a new tab

5. Cell culture and clinical sample validation

We selected TMEM59 as the focus of our next research step. Firstly, TMEM59-siRNA was transfected into LNCaP and PC3 cells to knock down TMEM59. The results indicated that the si-TMEM59 used in the experiment effectively suppressed the expression of TMEM59 (Fig. 5A). CCK8 and colony formation assays confirmed that si-TMEM59 significantly promoted the proliferation of LNCaP and PC3 cells (Fig. 5B, 5C). Transwell experiments showed that, compared to the control group, LNCaP and PC3 cells treated with si-TMEM59 exhibited higher invasive capabilities (Fig. 5D). This is consistent with our previous analysis, suggesting that downregulation of TMEM59 plays a role in promoting proliferation, migration, and invasion of prostate cancer cells. Subsequently, we further evaluated the expression of TMEM59 in normal and prostate cancer tissues. IHC results demonstrated that the protein expression level of TMEM59 in normal tissues was significantly higher than in tumor tissues (Fig. 5E).

Fig. 5. Cellular experiments and immunohistochemistry. (A) Reverse transcription–polymerase chain reaction (RT-PCR) shows the knockdown efficiency of TMEM59. (B) Cell counting kit-8 assay indicates enhanced LNCaP and PC3 cells proliferation after knockdown. (C) Colony formation assay in si-TMEM59 and control cells (×10 objective lens and ×10 eyepiece). (D) TMEM59-knockdown promotes prostate cancer cell metastasis in LNCaP and PC3 cells (×20 objective lens and ×10 eyepiece). (E) Immunohistochemistry of TMEM59 in tumor and normal tissues (×4 objective lens, ×10 eyepiece, and ×20 objective,×10 eyepiece). *p<0.05, **p<0.01, ***p<0.001.

Fig. 5

Open in a new tab

DISCUSSION

Our study, initiated from a clinical perspective, utilized clinical information for grouping in order to delineate subpopulations in the single-cell dataset. We employed hdWGCNA for modular gene selection, focusing on identifying cell subgroups with later clinical stages and stronger tumor invasiveness. Traditional bulk sequencing data analysis often relies on differential analysis at the gene selection step. In contrast, the advantage of single-cell sequencing lies in its cell-centric approach, providing a better understanding of cell-cell interactions within the tumor microenvironment [14]. Another crucial reason is that, compared to the comparison between tumors, which shows minimal differences, the comparison between tumors and normal tissues exhibits more variation. Therefore, our core research strategy involves utilizing the ultra-high resolution of single-cell sequencing to identify cell subgroups with significantly different expression between different clinical groups.

Meanwhile, in our analysis, we observed significant differences in cell communication among single-cell patients from different clinical subgroups. By reviewing existing studies, we found that these pathways play important roles in tumor progression. For instance, MK is associated with drug resistance mechanisms in various tumors and affects tumor invasion [15]. EGF, TWEAK, Wnt, and GDF have long been identified as playing critical roles in the development and progression of cancer cells, including but not limited to metastasis, invasion, and drug resistance [16,17]. In prostate cancer, these pathways also promote tumor progression. For example, EGF may enhance neuroendocrine-like differentiation in prostate cancer cells, increasing malignancy [18]. Activation of the Wnt signaling pathway may lead to bone metastasis and proliferation of prostate cancer cells [19]. Prostate cancer cells breach the prostatic capsule through enhanced invasiveness and adaptation to the smooth muscle barrier [3]. These findings suggest that the cell subpopulations we identified not only exhibit a higher number in invasive prostate cancer but may also display more harmful biological activities. Considering that the prostate capsule is primarily composed of smooth muscle, we speculate that this cell subtype is associated with the invasive development of prostate cancer. hdWGCNA is then employed for modular gene selection within the differential cell subgroups, enabling the discovery of differentially expressed genes among various clinical groups. These genes can be linked to clinical outcomes through the eQTL information in the GWAS database.

GWAS databases and single-cell sequencing are similar in scale, often involving statistical analyses with hundreds of thousands to millions of cases [20]. However, they lack precision in specific disease subdivisions and cannot precisely classify diseases, such as the TNM staging of tumors and specific survival information for patients. At the same time, we suggested that the causal relationships provided by GWAS data might be inaccurate. For example, WLS has been shown to promote neuroendocrine characteristics and proliferative capacity in neuroendocrine prostate cancer through Wnt signaling [21]. Additionally, it can enhance cell viability and resistance to enzalutamide in prostate cancer [22]. We also found that many MR results included only a single SNP, meaning we could only use the Wald ratio method for testing. However, this increases the probability of type II errors [23]. Therefore, we consider MR as an initial screening method. Following the MR analysis, we conducted gene selection analyses using TCGA and GEO data, which ultimately led to the identification of TMEM59. Pseudo-time analysis showed that TMEM59 is highly expressed in the early stages of tumor cell development and differentiation. According to MR results, TMEM59 is associated with the prostate, and combining second-generation sequencing data with patient clinical information confirmed that TMEM59 is a potentially protective factor for patients, as evidenced by the hazard ratio of TMEM59 being less than 1 for biochemical recurrence, indicating that its expression is a protective factor for patients experiencing biochemical recurrence.

TMEM59 is a membrane-associated protein located in the Golgi apparatus. Although the exact function of TMEM59 remains unclear, it has been demonstrated that TMEM59 protein expression inhibits the Golgi glycosylation of amyloid precursor protein (APP) and blocks the cleavage of APP by Golgi-resident α- and β-secretases, thereby inhibiting the formation of β-amyloid peptides that form amyloid plaques in Alzheimer’s disease [24]. Our experimental results suggest that the expression of TMEM59 may affect the proliferation and invasion capabilities of prostate cancer, and IHC results show that TMEM59 protein expression in normal tissues is significantly higher than in prostate cancer tissues. Since our study focused on differential cell subgroups rather than unique cell subgroups and the specificity between tumors was small, and IHC is a global indicator, we did not observe significant differences in TMEM59 in the IHC analysis of prostate tumor grouping. However, cell experiments demonstrated that downregulating TMEM59 enhances the proliferation and invasion of prostate cancer, providing a cellular perspective on the impact of TMEM59 on prostate cancer cells. Thus, TMEM59 may serve as a potential prognostic biomarker for prostate cancer patients, aiding in risk stratification and clinical decision-making.

In the existing research, there are limited studies on the association between TMEM59 and tumors, primarily focusing on its role in methylation and autophagy, with database analyses speculating its involvement in certain cancer occurrences. For example, TMEM59 has been found to be involved in the progression of gastric cancer in some studies [25]. Additionally, TMEM59 has been shown to enhance Wnt signaling by promoting the formation of the Wnt receptor signaling complex [26]. The Wnt pathway plays a crucial role in maintaining tissue homeostasis and serves as a key oncogenic driver in various cancer types [17]. Therefore, we hypothesize that the phenotypic effects of TMEM59 on the prostate may be mediated through the Wnt pathway, which will be the focus of our future research. In contrast, in other fields, TMEM59 has been found to play a role in autism, brain ischemic stroke, the maintenance and regeneration of olfactory sensory neurons, and olfactory perception. In these diseases, TMEM59 primarily exerts its effects through mechanisms such as autophagy and pyroptosis, participating in immune and inflammatory responses [27,28,29]. Our study is the first to investigate the relevance of TMEM59 to the occurrence and development of prostate cancer, combining multi-omics and experimental research to confirm that the low expression of TMEM59 may be associated with the progression of prostate cancer. However, our study has limitations, primarily in that it focuses on phenotypic observations without fully elucidating the underlying molecular mechanisms. Additionally, due to the high cost of single-cell sequencing and the limited availability of publicly accessible single-cell data at this stage, our study only included single-cell sequencing data from four patients. As sequencing technology advances, our future research will incorporate data from more patients. Moreover, the number of patients analyzed in IHC is also limited. Therefore, in our subsequent studies, we will conduct more in-depth upstream and downstream investigations on TMEM59, along with a larger-scale clinical study targeting TMEM59.

CONCLUSIONS

In this study, we utilized single-cell data in combination with hdWGCNA and MR to identify genes associated with prostate cancer. Further validation was performed using transcriptome data. Additionally, experimental evidence demonstrated that knockdown of TMEM59 may enhance the proliferation and invasion of LNCaP and PC3 cells, suggesting a potential therapeutic target for prostate cancer.

Acknowledgements

None.

Footnotes

Conflict of Interest: The authors have nothing to disclose.

Funding: This research was supported by Suzhou Clinical Medical Center for Urological Diseases (No. Szlcyxzx202106).

Author Contribution:
  • Conceptualization: XF, CF, WW.
  • Data curation: XF.
  • Formal analysis: XF, WW.
  • Funding acquisition: GC, ZT, MZ.
  • Investigation: XZ, KW.
  • Methodology: XF, CF, WW.
  • Project administration: GC, ZT, MZ.
  • Resources: GC, ZT, MZ.
  • Validation: All authors.
  • Writing – original draft: XZ, WW.
  • Writing – review & editing: All authors.

Data Sharing Statement

The authors confirm that the cell experiment data supporting the findings of this study are available within the article and its supplementary materials. The single cell sequencing data that support the findings of this study are openly available in GEO [https://www.ncbi.nlm.nih.gov/geo/], reference number [GSE193337] and [GSE116918].The GWAS summary data that support the findings of this study are openly available in IEU OPEN GWAS PROJECT at [https://gwas.mrcieu.ac.uk/], reference number [ukb-b-13348]. The bulk sequencing data and clinical information that support the findings of this study are openly available in TCGA (The Cancer Genome Atlas Program) at [https://portal.gdc.cancer.gov/], reference number [TCGAPRAD]. The analysis code file for the article can be accessed at the following URL: https://www.jianguoyun.com/p/DRqff_sQpLmNChjhw-QFIAA.

References

  • 1.Siegel DA, O'Neil ME, Richards TB, Dowling NF, Weir HK. Prostate cancer incidence and survival, by stage and race/ethnicity-United States, 2001-2017. MMWR Morb Mortal Wkly Rep. 2020;69:1473–1480. doi: 10.15585/mmwr.mm6941a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209–249. doi: 10.3322/caac.21660. [DOI] [PubMed] [Google Scholar]
  • 3.Schaeffer EM, Srinivas S, Adra N, An Y, Barocas D, Bitting R, et al. NCCN Guidelines® insights: prostate cancer, version 1.2023. J Natl Compr Canc Netw. 2022;20:1288–1298. doi: 10.6004/jnccn.2022.0063. [DOI] [PubMed] [Google Scholar]
  • 4.Achard V, Panje CM, Engeler D, Zilli T, Putora PM. Localized and locally advanced prostate cancer: treatment options. Oncology. 2021;99:413–421. doi: 10.1159/000513258. [DOI] [PubMed] [Google Scholar]
  • 5.Gillessen S, Armstrong A, Attard G, Beer TM, Beltran H, Bjartell A, et al. Management of patients with advanced prostate cancer: report from the advanced prostate cancer consensus conference 2021. Eur Urol. 2022;82:115–141. doi: 10.1016/j.eururo.2022.04.002. [DOI] [PubMed] [Google Scholar]
  • 6.Ren X, Zhang L, Zhang Y, Li Z, Siemers N, Zhang Z. Insights gained from single-cell analysis of immune cells in the tumor microenvironment. Annu Rev Immunol. 2021;39:583–609. doi: 10.1146/annurev-immunol-110519-071134. [DOI] [PubMed] [Google Scholar]
  • 7.Morabito S, Reese F, Rahimzadeh N, Miyoshi E, Swarup V. hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data. Cell Rep Methods. 2023;3:100498. doi: 10.1016/j.crmeth.2023.100498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Zuber V, Grinberg NF, Gill D, Manipur I, Slob EAW, Patel A, et al. Combining evidence from Mendelian randomization and colocalization: review and comparison of approaches. Am J Hum Genet. 2022;109:767–782. doi: 10.1016/j.ajhg.2022.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Cano-Gamez E, Trynka G. From GWAS to function: using functional genomics to identify the mechanisms underlying complex diseases. Front Genet. 2020;11:424. doi: 10.3389/fgene.2020.00424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Slovin S, Carissimo A, Panariello F, Grimaldi A, Bouché V, Gambardella G, et al. Single-cell RNA sequencing analysis: a step-by-step overview. Methods Mol Biol. 2021;2284:343–365. doi: 10.1007/978-1-0716-1307-8_19. [DOI] [PubMed] [Google Scholar]
  • 11.Hong B, Li Y, Yang R, Dai S, Zhan Y, Zhang WB, et al. Single-cell transcriptional profiling reveals heterogeneity and developmental trajectories of Ewing sarcoma. J Cancer Res Clin Oncol. 2022;148:3267–3280. doi: 10.1007/s00432-022-04073-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Jin S, Guerrero-Juarez CF, Zhang L, Chang I, Ramos R, Kuan CH, et al. Inference and analysis of cell-cell communication using CellChat. Nat Commun. 2021;12:1088. doi: 10.1038/s41467-021-21246-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Namekawa T, Ikeda K, Horie-Inoue K, Inoue S. Application of prostate cancer models for preclinical study: advantages and limitations of cell lines, patient-derived xenografts, and three-dimensional culture of patient-derived cells. Cells. 2019;8:74. doi: 10.3390/cells8010074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Li X, Wang CY. From bulk, single-cell to spatial RNA sequencing. Int J Oral Sci. 2021;13:36. doi: 10.1038/s41368-021-00146-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Saikia M, Cheung N, Singh AK, Kapoor V. Role of midkine in cancer drug resistance: regulators of its expression and its molecular targeting. Int J Mol Sci. 2023;24:8739. doi: 10.3390/ijms24108739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Fang L, Li F, Gu C. GDF-15: a multifunctional modulator and potential therapeutic target in cancer. Curr Pharm Des. 2019;25:654–662. doi: 10.2174/1381612825666190402101143. [DOI] [PubMed] [Google Scholar]
  • 17.Parsons MJ, Tammela T, Dow LE. WNT as a driver and dependency in cancer. Cancer Discov. 2021;11:2413–2429. doi: 10.1158/2159-8290.CD-21-0190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Cortés MA, Cariaga-Martinez AE, Lobo MV, Martín Orozco RM, Motiño O, Rodríguez-Ubreva FJ, et al. EGF promotes neuroendocrine-like differentiation of prostate cancer cells in the presence of LY294002 through increased ErbB2 expression independent of the phosphatidylinositol 3-kinase-AKT pathway. Carcinogenesis. 2012;33:1169–1177. doi: 10.1093/carcin/bgs139. [DOI] [PubMed] [Google Scholar]
  • 19.Murillo-Garzón V, Kypta R. WNT signalling in prostate cancer. Nat Rev Urol. 2017;14:683–696. doi: 10.1038/nrurol.2017.144. [DOI] [PubMed] [Google Scholar]
  • 20.Buniello A, MacArthur JAL, Cerezo M, Harris LW, Hayhurst J, Malangone C, et al. The NHGRI-EBI GWAS Catalog of published genome-wide association studies, targeted arrays and summary statistics 2019. Nucleic Acids Res. 2019;47:D1005–D1012. doi: 10.1093/nar/gky1120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Bland T, Wang J, Yin L, Pu T, Li J, Gao J, et al. WLS-Wnt signaling promotes neuroendocrine prostate cancer. iScience. 2021;24:101970. doi: 10.1016/j.isci.2020.101970. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Lombard AP, Liu C, Armstrong CM, D'Abronzo LS, Lou W, Evans CP, et al. Wntless promotes cellular viability and resistance to enzalutamide in castration-resistant prostate cancer cells. Am J Clin Exp Urol. 2019;7:203–214. [PMC free article] [PubMed] [Google Scholar]
  • 23.Bowden J, Holmes MV. Meta-analysis and Mendelian randomization: a review. Res Synth Methods. 2019;10:486–496. doi: 10.1002/jrsm.1346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Ullrich S, Münch A, Neumann S, Kremmer E, Tatzelt J, Lichtenthaler SF. The novel membrane protein TMEM59 modulates complex glycosylation, cell surface expression, and secretion of the amyloid precursor protein. J Biol Chem. 2010;285:20664–20674. doi: 10.1074/jbc.M109.055608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Wu J, Gu Y, Xiao Y, Xia C, Li H, Kang Y, et al. Characterization of DNA methylation associated gene regulatory networks during stomach cancer progression. Front Genet. 2019;9:711. doi: 10.3389/fgene.2018.00711. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Gerlach JP, Jordens I, Tauriello DVF, van't Land-Kuper I, Bugter JM, Noordstra I, et al. TMEM59 potentiates Wnt signaling by promoting signalosome formation. Proc Natl Acad Sci U S A. 2018;115:E3996–E4005. doi: 10.1073/pnas.1721321115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Meng J, Han L, Zheng N, Wang T, Xu H, Jiang Y, et al. Microglial Tmem59 deficiency impairs phagocytosis of synapse and leads to autism-like behaviors in mice. J Neurosci. 2022;42:4958–4979. doi: 10.1523/JNEUROSCI.1644-21.2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Zhang L, Wang T, Chen XF, Xu ZX, Cao JB, Sun H. TMEM59 protects against cerebral ischemic stroke by suppressing pyroptosis and microglial activation. Biochem Biophys Res Commun. 2021;543:72–79. doi: 10.1016/j.bbrc.2020.09.013. [DOI] [PubMed] [Google Scholar]
  • 29.Ma Z, Li W, Zhuang L, Wen T, Wang P, Yu H, et al. TMEM59 ablation leads to loss of olfactory sensory neurons and impairs olfactory functions via interaction with inflammation. Brain Behav Immun. 2023;111:151–168. doi: 10.1016/j.bbi.2023.04.005. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The authors confirm that the cell experiment data supporting the findings of this study are available within the article and its supplementary materials. The single cell sequencing data that support the findings of this study are openly available in GEO [https://www.ncbi.nlm.nih.gov/geo/], reference number [GSE193337] and [GSE116918].The GWAS summary data that support the findings of this study are openly available in IEU OPEN GWAS PROJECT at [https://gwas.mrcieu.ac.uk/], reference number [ukb-b-13348]. The bulk sequencing data and clinical information that support the findings of this study are openly available in TCGA (The Cancer Genome Atlas Program) at [https://portal.gdc.cancer.gov/], reference number [TCGAPRAD]. The analysis code file for the article can be accessed at the following URL: https://www.jianguoyun.com/p/DRqff_sQpLmNChjhw-QFIAA.

Articles from The World Journal of Men's Health are provided here courtesy of Korean Society for Sexual Medicine and Andrology

RESOURCES