|Ahead of print publication
Genome-wide association analysis reveals regulation of at-risk loci by DNA methylation in prostate cancer
Qiang Liu1,2, Gang Liu3, Darryl T Martin2, Yu-Tong Xing4, Robert M Weiss2, Jun Qi1, Jian Kang1
1 Department of Urology, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai 200092, China
2 Department of Urology, Yale University School of Medicine, New Haven, CT 06510, USA
3 Key Lab of Systems Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China
4 Institute of Estuarine and Coastal Research, East China Normal University, Shanghai 200062, China
|Date of Submission||18-Nov-2020|
|Date of Acceptance||21-Jan-2021|
|Date of Web Publication||23-Mar-2021|
Department of Urology, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai 200092
Department of Urology, Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, Shanghai 200092
Source of Support: None, Conflict of Interest: None
Epigenetic changes are potentially important for the ontogeny and progression of tumors but are not usually studied because of the complexity of analyzing transcript regulation resulting from epigenetic alterations. Prostate cancer (PCa) is characterized by variable clinical manifestations and frequently unpredictable outcomes. We performed an expression quantitative trait loci (eQTL) analysis to identify the genomic regions that regulate gene expression in PCa and identified a relationship between DNA methylation and clinical information. Using multi-level information published in The Cancer Genome Atlas, we performed eQTL-based analyses on DNA methylation and gene expression. To better interpret these data, we correlated loci and clinical indexes to identify the important loci for both PCa development and progression. Our data demonstrated that although only a small proportion of genes are regulated via DNA methylation in PCa, these genes are enriched in important cancer-related groups. In addition, single nucleotide polymorphism analysis identified the locations of CpG sites and genes within at-risk loci, including the 19q13.2–q13.43 and 16q22.2–q23.1 loci. Further, an epigenetic association study of clinical indexes detected risk loci and pyrosequencing for site validation. Although DNA methylation-regulated genes across PCa samples are a small proportion, the associated genes play important roles in PCa carcinogenesis.
Keywords: CpG sites; DNA methylation; expression quantitative trait loci; genome-wide association study; prostate cancer
Article in PDF
|How to cite this URL:|
Liu Q, Liu G, Martin DT, Xing YT, Weiss RM, Qi J, Kang J. Genome-wide association analysis reveals regulation of at-risk loci by DNA methylation in prostate cancer. Asian J Androl [Epub ahead of print] [cited 2021 Aug 5]. Available from: https://www.ajandrology.com/preprintarticle.asp?id=311757
Qiang Liu, Gang Liu
These authors contributed equally to this work.
| Introduction|| |
Prostate cancer (PCa) is a common malignancy and leading cause of cancer death among men in the US where it is estimated that approximately 191 930 new PCa cases are expected to be diagnosed in 2020. Further, according to a Chinese statistical report, 60 300 PCa cases were confirmed in 72 population-based cancer registries in China from 2009 to 2011, with a mortality of 26 600, resulting in rankings of sixth and eleventh place, respectively, among all cancers affecting men, in China. As the Chinese population continues to age, both the incidence and mortality of PCa are expected to continue increasing over the next two decades. PCa cells are known to harbor a variety of genetic defects, including gene mutations and translocations, all of which provide the cells with new capabilities for dysregulated proliferation, immune system evasion, tissue invasion and destruction, inappropriate survival, and metastasis. Furthermore, there is abundant evidence that along with genetic changes, somatic epigenetic alterations also contribute to PCa carcinogenesis and metastasis.,,, Epigenetic gene inactivation in cancer cells is largely based on transcriptional silencing mediated by the aberrant CpG methylation of CpG-rich promoter regions.,,,, DNA methylation is a widely recognized epigenetic marker associated with diagnosis and prognosis in many malignancies.,, Notably, abnormal DNA methylation has been reported to contribute to the occurrence and progression of PCa., Previous studies of DNA methylation and PCa risk have found that specific promoter sequences are hypermethylated at a higher frequency in PCa tumor tissues than those in nontumor tissues.,,,
Quantitative genetics has made a significant progress in revealing the genetic bases of complex traits, especially in developing sophisticated tools to identify the location of genes that impact complex traits. A region of the genome contributing to the variation in a quantitative trait, also known as quantitative trait loci (QTLs), has been used to study gene expression phenotypes (expression quantitative trait loci [eQTLs]) on a massive scale. DNA methylation quantitative trait loci (meQTLs) have been identified in pathological and physiological contexts. The genome-wide gene expression studies can provide information on genetic variation that affects gene expression levels. We can use linkage or association mapping to map cis- and trans-acting factors for many genes to explain the inheritance patterns. Previous reports showed that cis/trans-meQTLs could target different CpGs and clarified DNA methylation involvement in diseases and cancers.,,,, Thus, eQTL analysis is a straightforward and popular method for discovering regulatory genome sites,, and detecting underlying associations across the genome in PCa studies.,
The Cancer Genome Atlas (TCGA, http://cancergenome.nih.gov/) is a database that collects multiple types of “omics” from thousands of samples and provides public data access to researchers. Accordingly, modified eQTL methods involving calculations of correlations among gene expression, phenotype, DNA methylation, copy number variations, and single-nucleotide polymorphisms (SNPs) have proven to be a powerful tool.,, However, genome-wide correlations between gene expression, DNA methylation, and clinical phenotypes in PCa are not yet understood.
Therefore, we performed a meQTL analysis of PCa samples from the TCGA database. Although genes regulated via methylation comprise a minor proportion of the genome (0.8%), these genes are enriched in some gene ontology (GO) groups and in important canonical cancer-related pathways. In our study, we mainly identified meQTL pairs on chromosomes 16q and 19q, which have been reported as high-risk regions for PCa, using SNP analyses. We also identified DNA methylation regions and genes associated with clinical indexes at 11q13 and 16q13 and found several genes regulated by DNA methylation that are important for prognosis. According to our meQTL analyses, we selected some androgen receptor (AR) gene-related CpG sites and some sites that are altered during PCa genesis, and correlated them with Gleason score.
| Materials and Methods|| |
A transcriptome of a prostate adenocarcinoma (PRAD) gene evaluation was downloaded with the TCGA dataset. The nonprimary PCa samples were discarded. Furthermore, the primary tumor samples not providing either gene expression (evaluated with RNA sequencing [RNA-seq]) or DNA methylation information also were excluded from later analysis. As result, a total of 419 samples were used for meQTL identification. Methylation data also were downloaded for the TCGA dataset and were correlated with expression data. Methylation levels at the evaluated sites were estimated using an Illumina Human Infilium 450k BeadChip (Illumina, San Diego, CA, USA). After sample normalization, samples were combined to a methylation matrix according to the gene ID.
We used R software MatrixeQTL package (version 2.15.1; R Project for Statistical Computing, Vienna, Austria) for eQTL analysis. Each methylated CpG level was regarded as a continuous variable rather than a discrete variable. Correlations between the methylation level of each CpG site and each gene were evaluated using MatrixeQTL. As DNA methylation mostly influences the expression of genes via promoter regions, we distinguished cis- and trans-regulation for further analysis. The gene and methylation CpG sites were extracted from Illumina Human Infilium 450k BeadChip annotation files and University of California, Santa Cruz (UCSC) reference gene list locations. For reference genes, the gene location was counted from the transcription start site to terminal site. We set a P value threshold of 1 × 10−5 for cis-regulated and 1 × 10−6 for trans-regulated eQTLs. Cis-regulated meQTLs were defined as interacting pairs with tested DNA methylation sites and genes <1 megabase (MB), whereas trans-regulated meQTLs were defined as pairs with genes >1 MB or located on other chromosomes.
Tissue specimens and bisulfite modification of DNA
A total of 70 PCa patients who underwent radical prostatectomy at Xinhua Hospital affiliated to Shanghai Jiao Tong University School of Medicine (Shanghai, China), between July 2014 and January 2017, were enrolled in this study. Informed consent was obtained from all subjects. All cases were histologically confirmed and had clinical stage II or III tumors, no clinical evidence of lymph node or distant metastasis, available pathology specimens, and complete clinical and serum prostate-specific antigen (PSA) data. Patients with missing any variable needed to accurately assign them to a risk group (PSA, tumor [T] stage, or Gleason score) were excluded. Patients with missing information on any available demographic variables were excluded. All samples were retrieved from the archive of the Institute of Pathology, Xinhua Hospital, and were anonymously analyzed in accordance with the guidelines of the Ethics Committee of Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine. We confirm that all experimental protocols were approved by the local ethics committee (approval No. XHEC-D-2016-005).
Patients were categorized as low-, moderate-, or high-risk PCa based on the 2015 National Comprehensive Cancer Network guidelines [Table 1]. All 70 cases included tumor tissues and adjacent nontumor tissues. Prostate tissue samples from these cases were obtained at the time of radical prostatectomy, and tumor cell contents were determined to exceed 70.0% of all tissues. Tissue microdissection yielded tumor and adjacent nontumor tissues from which DNA was extracted using an FFPE DNA Kit (Omega Bio-Tek, Norcross, GA, USA) according to the manufacturer's instructions. Extracted DNA concentrations were measured using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA) and subsequently adjusted to approximately 40 ng ml−1. Prepared DNA was subjected to bisulfite treatment using the EZ DNA Methylation-Gold kit (ZYMO, Orange, CA, USA); converted DNA was dissolved in Tris-EDTA (TE) buffer and stored at or below −20°C for later use.
Primer design and PCR
Specific primers were designed for 30 loci using the publicly available MethPrimer software package (http://www.urogene.org/methprimer/). A complete list of the primer pairs is available in [Supplementary Table 1 [Additional file 1]]. Biotinylated reverse primers were substituted with 5′-tailed unlabeled reverse primers (aaccttcaacaccccaaccatata), allowing single (expansive) biotinylated primers to be used for subsequent pyrosequencing™. All primers and tag sequences were provided by Sangon (Shanghai, China).
PCR amplification of DNA was conducted using a nested-PCR protocol, using the primers shown in [Supplementary Table 1]. Two rounds of amplification reactions were performed. In the first round, a total reaction volume of 10 μl contains 1 mmol l−1 first primers (1 μl), bisulfite converted DNA (1 μl), ×2 PCR master mix (5 μl), and ddH2O (2 μl). In the second round, the total reaction volume was 30 μl: 4 μl of first-round PCR product as template, plus 10 mmol l−1 second primers (0.5 μl), ×2 PCR Master Mix (15 μl), and ddH2O (10 μl). PCR conditions were as follows: 94°C for 5 min, 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 40 s, elongation at 72°C for 30 s, and an additional elongation step at 72°C for 7 min. Approximately 8 μl of each PCR product was separated by electrophoresis using a 3% agarose gel stained with GelRed (Invitrogen, Carlsbad, CA, USA) for 40 min at 120 V and visualized using a Gene Genius bio-imaging system (Syngene, Cambridge, UK).
Pyrosequencing™ methylation analysis
PCR products (20 μl) were added to a mix comprising Streptavidin Sepharose HP™ (3 μl; GE Healthcare, Dornstadt, Germany) and binding buffer (37 μl; Qiagen, Hilden, Germany). The contents were mixed at 2000g (Centrifuge 5810R, Eppendorf, Hamburg, Germany) for 10 min at room temperature. Using the Vacuum Prep Tool™ (Qiagen), according to the manufacturer's instructions, single-stranded PCR products were prepared. Sepharose beads with attached single-stranded templates were released into a PSQ 96 Plate Low™ (Qiagen) containing a mix of 40 μl annealing buffer (Qiagen) and the corresponding sequencing primer at 400 nmol l−1 [Supplementary Table 1]. Pyrosequencing™ reactions were performed in a PyroMark ID System (Qiagen), according to the manufacturer's instructions, using the PyroMark Gold 96 Reagent Kit (Qiagen). CpG site quantification was performed using Pyro Q-CpG™ methylation software (Qiagen).
The cis- and trans-meQTL sites were identified using R package “MatrixeQTL”. The P values were adjusted using Bonferroni method, and cis-meQTLs with false discovery rate (FDR) <1 × 10−5 and trans-meQTLs with FDR <1 × 10−6 were identified as meQTLs. Association significance between clinical indexes and gene expression/methylation levels was identified using analysis of variance (ANOVA) test (P < 0.05 being considered as significant). Gene expression comparison between recurrence and nonrecurrence group was performed using Student's t-test, and P < 0.05 was considered as statistically significant.
| Results|| |
Methylated meQTLs in PCa
We subjected PCa samples including 419 samples from TCGA to a meQTL analysis using expression levels evaluated via RNA-seq and CpG site methylation levels in CpG island (CGI) regions. Among 20 321 genes, the methylation levels of 485 513 CpG loci were used for meQTL identification. We identified 5852 cis-regulation and 5 156 662 trans-regulation meQTL pairs in our dataset. Cis-regulation pairs included 1717 genes (8.4% of all genes tested) and 4895 corresponding CpG sites, with an FDR (Bonferroni) of <0.01. Among these pairs, 784 genes (45.7%) were regulated by multiple CpG sites, and 1661 CpG sites (33.9%) regulated multiple genes. We noticed that adenosine triphosphate (ATP)-binding cassette subfamily A member 17, pseudogene (ABCA17P), chromosome 11 open reading frame 85 (C11orf85), and cyclic adenosine monophosphate (cAMP) responsive element binding protein 3 like 3 (CREB3L3) were regulated according to the methylation levels of 135, 95, and 93 nearby CpG sites, respectively, which are the top three CpG sites-related genes, suggesting the importance of DNA methylation on these genes. We also found that significantly high numbers of CpG sites in meQTLs were located on 19q13.2–q13.43 (P = 1 × 10−11) and 16q22.2–q23.1 (P = 1 × 10−11), as shown in [Figure 1] and [Supplementary Table 2 [Additional file 2]], consistent with previous reports., According to GATHER (http://gather.genome.duke.edu/), the genes with significant involvement are located at 19q13, 16p13, 11q13, and 17q25 [Supplementary Table 3 [Additional file 3]]. Through a GO and pathway analysis using the database for annotation, visualization, and integration discovery (DAVID), we discovered that although most GO groups appeared stochastic, several signaling pathways associated with PCa, including the Janus kinase/signal transducer and activator of tran-ions (JAK/STAT), vascular endothelial growth factor (VEGF), and cytokine–cytokine interactions, also are involved [Supplementary Table 4 [Additional file 4]] and [Supplementary Table 5 [Additional file 5]]. Finally, the analysis of trans-regulation pairs identified 18 999 genes (93.5% of tested genes) and 239 808 CpG sites, with an FDR of <1 × 10−6. In summary, SNP analysis identified the locations of CpG sites and genes within at-risk loci, including the 19q13.2–q13.43 and 16q22.2–q23.1 loci.
|Figure 1: Circos plot showing the CpG density of eQTLs. The density of CpG sites associated with eQTLs (from inside to outside) across the genome. Each wedge represents a chromosome and different colors show different cytobands. In the inner track, from inner to outer, the fold line represents the values. Higher values mean that more meQTL-related CpG sites are over-present in these regions. eQTLs: expression quantitative trait loci; meQTL: methylation quantitative trait loci.|
Click here to view
Gene/CpG methylation and clinical information
To identify the mechanism by which these CpG sites affect carcinogenesis and development, we performed a correlation analysis and an ANOVA between gene expression levels and clinical indexes. The evaluated genes included 20 532 genes from the TCGA database that were evaluated via RNA-seq and correlated with clinical information, including primary and secondary Gleason score, node invasion stage, biochemical recurrence indicators, most recent PSA level, and clinical and pathological primary tumor stages. Each clinical index was associated with a number of genes, as shown in [Table 2].
|Table 2: Number of genes identified significantly associated with clinical observations|
Click here to view
Genes associated with the same clinical information may affect each other. We further performed a similar analysis to compare methylation sites and clinical information and identified 57 719 CpG sites significantly correlated with the aforementioned clinical information [Table 3].
|Table 3: Number of CpG sites significantly associated with clinical observations|
Click here to view
Methylation sites influence clinical performance by regulating gene expression
A canonical function of DNA methylation is the blockade of transcription factors from binding gene regulatory elements, thus inhibiting gene expression. To reduce the FDR and determine causality, we sought CpG sites that were significantly correlated with gene expression via cis-regulation according to methylated QTLs. These genes were positively correlated further with clinical information [Supplementary Table 6 [Additional file 6]], and the CpG sites also were positively correlated with the same clinical indexes. We identified 92 paired methylated QTLs that fulfilled the aforementioned criteria. Most methylated QTLs were associated with Gleason score; a DAVID-based GO analysis revealed that these genes were significantly enriched for DNA repair and cell cytoskeleton. Most of the sites associated with Gleason score were located in 11q13 and 16q13, and their correlated genes included essential meiotic structure-specific endonuclease subunit 2 (EME2), potassium channel tetramerization domain containing 13 (KCTD13), kelch like family member 17 (KLHL17), and WD repeat domain 90 (WDR90). Of these, EME2 is associated with genomic stability maintenance, whereas reports of the other genes in the context of all types of cancer are scarce.
We identified that calcium/calmodulin-dependent serine protein kinase interacting protein 1 (CASKIN1) was enriched with CpGs and CASKIN1 itself was also significantly associated with genes related to biochemical cancer recurrence (P = 0.001; [Figure 2]). According to the methylated QTL algorithm, these CpG sites' methylation levels may act as cis-regulators of CASKIN1, suggesting that the methylation of nearby CpG sites probably influences the expression of CASKIN1 and thus affects the recurrence in these cases.
|Figure 2: The violin plot of relative expression of CASKIN1 between biomedical recurrence and nonrecurrence subgroup evaluated by TCGA. The expression of CASKIN1 is significantly different between recurrence and nonrecurrence group (P = 0.001). CASKIN1: calcium/calmodulin-dependent serine protein kinase interacting protein 1; TCGA: The Cancer Genome Atlas.|
Click here to view
Methylated QTLs in PCa-related genes
An analysis of these regulation pairs should include the regulatory status in both prostate and PCa cells. To extract useful information for later analysis, we selected meQTLs concerning genes associated with PCa genesis and metastasis.,,,,, Ninety-two genes were used for meQTL selection. In our meQTL dataset, 38 cis-regulated meQTLs existed near these sites, and 1753 trans-regulated eQTLs were identified. Genes of interest among cis-eQTLs included alanyl aminopeptidase, membrane (ANPEP), activating transcription factor 3 (ATF3), B cell leukemia/lymphoma 2 (BCL2) associated X (BAX), and early growth response 1/3 (EGR1/3), suggesting that the expression levels of these PCa-related genes are regulated by the methylation levels of nearby CpG sites. For further analysis, these gene and methylation findings were validated in 70 pairs of clinical tumorous and nontumorous tissue samples. The patients from whom samples were collected were classified as low, moderate, and high risk according to Gleason score, mean PSA, and tumor/node/metastasis (TNM) clinical staging; a variety of tissues from different risk levels were subjected to methylation quantification via pyrosequencing™.
In our results, the false discovery rate was relatively high because we did not detect the expression levels of the selected meQTL-related genes. However, we still detected several differences in methylation sites between normal and tumorous samples [Figure 3]. These sites included meQTL pairs involving FosB proto-oncogene, activator protein-1 (AP-1) transcription factor subunit (FOSB), ANPEP, and ras-related dexamethasone-induced 1 (RASD1). In tumorous samples, the methylation levels at these sites, particularly cg20664996 near RASD1, were low. We also noticed that in high- and moderate-risk samples, the methylation levels at this site reached 0. RASD1 has been reported as an apoptosis- and ras-related gene that prevents aberrant cell growth in several cell lines. In short, an epigenetic association study of clinical indexes detected risk loci and pyrosequencing for site validation. Although the number of DNA methylation-regulated genes across PCa samples is a small proportion, the associated genes play essential roles in PCa carcinogenesis.
|Figure 3: Four CpG loci that were significantly differently methylated according to the validated dataset. In each CpG locus, the left dots indicate the relative methylation level of the normal tissues and the right is the corresponding cancerous tissues, the P values between the normal tissues and cancerous tissues are shown from left to right.|
Click here to view
| Discussion|| |
DNA methylation is one among the most common and well-characterized epigenetic changes in PCa.,,,,,, The importance of large-scale methylome studies has increased in biomedical fields, and high-throughput sequencing technologies promise sensitive, quantitative, and high-resolution large-scale DNA methylation analyses. A genome-wide association study (GWAS) based on QTL analysis is an effective method with which cancer-related genes and risk loci throughout the genome can be studied;,, similarly, the usefulness of methylation-based GWAS has been reported. However, despite reports on the association of DNA methylation with carcinogenesis and development in the prostate,,,, genome-wide analyses of the associations between gene expression, DNA methylation, and clinical information related to PCa continue to yield vague results. Here, we utilized TCGA to obtain DNA methylation data from more than 400 000 CpG sites, expression levels of over 20 000 genes, and clinical indexes for 419 PCa samples, to perform a methylated QTL analysis in which we identified 5852 cis-regulating pairs comprising 1712 genes. Compared with previous breast cancer studies, only 0.8% of genes associated with PCa are affected by DNA methylation, and these genes exhibit significant involvement in several pathways that previously have been reported to play important roles in PCa genesis, including the JAK/STAT and VEGF pathways., We also noticed that the associated genes and locations were significantly enriched at several genome locations, including 19q13.2–q13.43 and 16q22.2–q23.1. As these are cis-interactions within 1 MB, the regulated genes also are enriched at these sites. Previous reports of SNPs and PCa noted an association between SNPs at 19q13 and PCa risk. Another GWAS of a Chinese population demonstrated 19q13 as a novel risk locus for PCa, and a decade-long study detected chromosomal deletion and gene suppression at 16q22 in PCa. It was worth noting that this region was reported to show a loss of heterogeneity in breast cancer. In addition to somatic mutations, SNPs, and copy number variations, our data suggest that DNA methylation may also influence PCa genesis and probably plays important roles in this process by regulating the transcription of this gene nexus.
Although a previous study based on eQTL algorithms aimed to define the relationships between genomic, epigenomic, and transcriptomic changes,, the link between the data obtained and clinical information remains unclear. Heyn et al. reported a TCGA QTL analysis using 13 tumor types from TCGA, including PCa, but their study lacks correlations to PCa-specific clinical variables such as Gleason score and PSA. Their data sets (cancer samples and adjacent normal tissues) did not include control subjects without the disease; future studies are needed to determine if GWAS risk alleles exhibit similar relationships in cancer-unrelated donors.
To further investigate the relationships of DNA methylation and gene expression with clinical information, we performed an association study between the three pairs (expression-methylation, methylation-clinical information, and clinical information-expression) and identified 92 genes regulated by CpG site methylation. Furthermore, we associated the expression levels of these genes with clinical information, particularly the Gleason score. GO analysis revealed that these 92 genes were significantly enriched with respect to DNA repair and cell cytoskeleton, suggesting that the genes probably direct the development and prognosis of PCa. Of these genes, we noted that EME2 previously was reported to maintain genomic stability, whereas the roles of the other genes were poorly elucidated. A locus analysis of these 92 genes demonstrated significant enrichment at 11q13 (P = 0.008, 3 genes) and 16q13 (P < 0.0001, 6 genes). The genes identified were enriched according to an in-house R script of the hypergeometric distribution, using the genes assayed for meQTLs as background. Enriched genomic loci with P < 0.01 were identified, and some genes associated with the mQTL were significantly enriched for DNA repair and cell cytoskeleton. Most of the sites associated with Gleason score were in 11q13 and 16q13, and their correlated genes included EME2, KCTD13, KLHL17, and WDR90. These special genes played critical roles in the biological and pathological characteristics of an organism.
Studies based on SNPs or somatic mutation profiling have shown that 11q13 and 16q13 loci are associated with aggressiveness, invasion, and poor prognosis of PCa. Our results show that epigenetic alterations in these regions are associated with genes considered necessary for the progression of PCa and suggest that methylation at 11q13 and 16q13 loci is essential to PCa progression. We were the first to prove that CASKIN1 was enriched with CpGs, and CASKIN1 itself was also significantly associated with genes related to biochemical cancer recurrence. Although the mechanism of how CASKIN1 affects PCa is unclear, its interacting protein, CASK has been associated with survival of patients with colorectal cancer.
For validation, we collected 70 clinical tissue samples from 2014 to 2017. Although these patients are Asian, we used the 2020 National Comprehensive Cancer Network (NCCN) guidelines as a criterion to classify distinct groups. Some studies suggested racial differences in PCa. Non-Whites were associated with a significantly higher likelihood of presenting with high-risk PCa (22.9% of Hispanics, 23.8% of Blacks, and 23.3% of those of other races compared with 19.0% of Whites; all P < 0.001). Age-adjusted PCa mortality in black men was more than double that of white men (42.0/100 000 vs 18.7/100 000). Asian men had a lower incidence of PCa and death than white men. Chinese and Japanese men have a relatively low risk of PCa than their European and North American counterparts.
The method we describe is based on detecting methylation and expression differences between samples of PCa. Therefore, it aims to identify correlations that occur within particular subsets of cases. For example, we found that some genes of interest among cis-eQTLs have played essential roles in PCa, including ANPEP,, ATF3,, BAX, and EGR1/3. These genes are involved in tumorigenesis, metastasis, and recurrence of PCa.
We selected 30 high-confidence CpG sites that have been reported to associate with PCa-related genes and evaluated methylation at related CpG sites [Supplementary Table 7 [Additional file 7]]. Despite a relatively high FDR, we determined differences in the methylation levels of several CpG sites between cancerous and noncancerous tissues (associated genes included ANPEP and FOSB). FDR is a tremendous concern for this study. To reduce FDR, we first enrolled the samples using the inclusion criteria described in the methods part. Second, we selected the P-value of meQTL pairs as 1 × 10−5 instead of 0.01/0.05. Last, the FDR values were calculated using Bonferroni, which is the most rigorous method available, and 0.01 was used for cutoff.
We further identified a correlation between the Gleason score and methylation level at cg20664996, a CpG location associated with RASD1 expression, according to an eQTL analysis. RASD1 is a 30 kDa G-protein that belongs to the Ras superfamily of small GTPases. Several robustly upregulated bicalutamide-dependent genes were identified by micro-array in LNCaP-ARW741L cells that were not significantly enhanced by dihydrotestosterone (DHT) in the LNCaP-LacZ line, including RASD1. Liu et al. showed that formononetin inhibited cell proliferation and induced apoptosis in DU-145 cells throughout the RASD1/mitogen-activated protein kinase (MAPK)/BAX pathway in PCa. The selected high-confidence CpG sites are reported to associate with some PCa-related genes, including FOSB, secretoglobin family 1A member 1 (SCGB1A1), mucin 16 (MUC16), alpha-methylacyl-coenzyme A racemase (AMACR), and glycine N-methyltransferase (GNMT). Emerging evidence has shown that these genes may be useful diagnostic and prognostic biomarkers for PCa.
| Conclusions|| |
We demonstrated that novel meQTL pairs were associated with PCa, and for the first time, we identified the locations of CpG sites and genes within at-risk loci, including the 19q13.2–q13.43 and 16q22.2–q23.1 loci. We further used pyrosequencing™ for validation and identified several genes that may impact the development and prognosis of PCa.
| Author Contributions|| |
QL conceived and participated in its design, searched databases, and extracted and assessed studies. QL and GL carried out the statistical analysis and interpretation of data and drafted the main manuscript. QL and DTM prepared the tables and figures. QL, YTX, and RMW prepared the figures and supplementary tables. QL, JK, and JQ participated in the conceptualization and design of the manuscript, performed the selection of studies, and drafted the manuscript. All authors reviewed and approved the final manuscript.
| Competing Interests|| |
All authors declare no competing interests.
| Acknowledgments|| |
We thank all the study participants. This study was supported by the Projects of National Science Foundation of China (No. 81070600 and 81570684) and Projects of the Shanghai Committee of Science and Technology, China (No. 14430720800, 134119a0600, and 11ZR1424100).
Supplementary Information is linked to the online version of the paper on the Asian Journal of Andrology website.
| References|| |
- Barrett CS, Millena AC, Khan SA. TGF-β effects on prostate cancer cell migration and invasion require FosB. Prostate 2017; 77: 72–81.
- Febbo PG, Mulligan MG, Slonina DA, Stegmaier K, Di Vizio D, et al. Literature Lab: a method of automated literature interrogation to infer biology from microarray analysis. BMC Genomics 2007; 8: 461.
- Shijubo N, Kawabata I, Sato N, Itoh Y. Clinical aspects of Clara cell 10-kDa protein/uteroglobin (secretoglobin 1A1). Curr Pharm Des 2003; 9: 1139–49.
- Ernst T, Hergenhahn M, Kenzelmann M, Cohen CD, Bonrouhi M, et al. Decrease and gain of gene expression are equally discriminatory markers for prostate carcinoma: a gene expression analysis on total and microdissected prostate tissue. Am J Pathol 2002; 160: 2169–80.
- Aithal A, Rauth S, Kshirsagar P, Shah A, Lakshmanan I, et al. MUC16 as a novel target for cancer therapy. Expert Opin Ther Targets 2018; 22: 675–86.
- Dharma Rao T, Park KJ, Smith-Jones P, Iasonos A, Linkov I, et al. Novel monoclonal antibodies against the proximal (carboxy-terminal) portions of MUC16. Appl Immunohistochem Mol Morphol 2010; 18: 462–72.
- Culig Z, Puhr M. Interleukin-6 and prostate cancer: current developments and unsolved questions. Mol Cell Endocrinol 2018; 462: 25–30.
- Baron VT, Pio R, Jia Z, Mercola D. Early Growth Response 3 regulates genes of inflammation and directly activates IL6 and IL8 expression in prostate cancer. Br J Cancer 2015; 112: 755–64.
- Qiu Y, Ravi L, Kung HJ. Requirement of ErbB2 for signalling by interleukin-6 in prostate carcinoma cells. Nature 1998; 393: 83–5.
- Vinarskaja A, Yamanaka M, Ingenwerth M, Schulz WA. DNA methylation and the HOXC6 paradox in prostate cancer. Cancers (Basel) 2011; 3: 3714–25.
- Costa VL, Henrique R, Ribeiro FR, Carvalho JR, Oliveira J, et al. Epigenetic regulation of Wnt signaling pathway in urological cancer. Epigenetics 2010; 5: 343–51.
- Yee DS, Tang Y, Li X, Liu Z, Guo Y, et al. The Wnt inhibitory factor 1 restoration in prostate cancer cells was associated with reduced tumor growth, decreased capacity of cell migration and invasion and a reversal of epithelial to mesenchymal transition. Mol Cancer 2010; 9: 162.
- Wissmann C, Wild PJ, Kaiser S, Roepcke S, Stoehr R, et al. WIF1, a component of the Wnt pathway, is down-regulated in prostate, breast, lung, and bladder cancer. J Pathol 2003; 201: 204–12.
- Göhler S, Da Silva Filho MI, Johansson R, Enquist-Olsson K, Henriksson R, et al. Functional germline variants in driver genes of breast cancer. Cancer Causes Control 2017; 28: 259–71.
- Farkas SA, Vymetalkova V, Vodickova L, Vodicka P, Nilsson TK. DNA methylation changes in genes frequently mutated in sporadic colorectal cancer and in the DNA repair and Wnt/β-catenin signaling pathway genes. Epigenomics 2014; 6: 179–91.
- Kim N, Hong Y, Kwon D, Yoon S. Somatic mutaome profile in human cancer tissues. Genomics Inform 2013; 11: 239–44.
- Unno K, Roh M, Yoo YA, Al-Shraideh Y, Wang L, et al. Modeling African American prostate adenocarcinoma by inducing defined genetic alterations in organoids. Oncotarget 2017; 8: 51264–76.
- Rastogi A, Ali A, Tan SH, Banerjee S, Chen Y, et al. Autoantibodies against oncogenic ERG protein in prostate cancer: potential use in diagnosis and prognosis in a panel with C-MYC, AMACR and HERV-K Gag. Genes Cancer 2016; 7: 394–413.
- Hessels D, Schalken JA. Urinary biomarkers for prostate cancer: a review. Asian J Androl 2013; 15: 333–9.
- Chen M, Huang YL, Huang YC, Shui IM, Giovannucci E, et al. Genetic polymorphisms of the glycine N-methyltransferase and prostate cancer risk in the health professionals follow-up study. PLoS One 2014; 9: e94683.
- Ianni M, Porcellini E, Carbone I, Potenzoni M, Pieri AM, et al. Genetic factors regulating inflammation and DNA methylation associated with prostate cancer. Prostate Cancer Prostatic Dis 2013; 16: 56–61.
- Ottaviani S, Brooke GN, O'Hanlon-Brown C, Waxman J, Ali S, et al. Characterisation of the androgen regulation of glycine N-methyltransferase in prostate cancer cells. J Mol Endocrinol 2013; 51: 301–12.
- Song YH, Shiota M, Kuroiwa K, Naito S, Oda Y. The important role of glycine N-methyltransferase in the carcinogenesis and progression of prostate cancer. Mod Pathol 2011; 24: 1272–80.
- O'Neill D, Jones D, Wade M, Grey J, Nakjang S, et al. Development and exploitation of a novel mutant androgen receptor modelling strategy to identify new targets for advanced prostate cancer therapy. Oncotarget 2015; 6: 26029–40.
- Liu XJ, Li YQ, Chen QY, Xiao SJ, Zeng SE. Up-regulating of RASD1 and apoptosis of DU-145 human prostate cancer cells induced by formononetin in vitro. Asian Pac J Cancer Prev 2014; 15: 2835–9.
- Sørensen KD, Abildgaard MO, Haldrup C, Ulhøi BP, Kristensen H, et al. Prognostic significance of aberrantly silenced ANPEP expression in prostate cancer. Br J Cancer 2013; 108: 420–8.
- Larkin SE, Holmes S, Cree IA, Walker T, Basketter V, et al. Identification of markers of prostate cancer progression using candidate gene expression. Br J Cancer 2012; 106: 157–65.
- Sharad S, Srivastava A, Ravulapalli S, Parker P, Chen Y, et al. Prostate cancer gene expression signature of patients with high body mass index. Prostate Cancer Prostatic Dis 2011; 14: 22–9.
- Avvakumova S, Galbiati E, Sironi L, Locarno SA, Gambini L, et al. Theranostic nanocages for imaging and photothermal therapy of prostate cancer cells by active targeting of neuropeptide-Y receptor. Bioconjug Chem 2016; 27: 2911–22.
- Ueda K, Tatsuguchi A, Saichi N, Toyama A, Tamura K, et al. Plasma low-molecular-weight proteome profiling identified neuropeptide-Y as a prostate cancer biomarker polypeptide. J Proteome Res 2013; 12: 4497–506.
- Ruscica M, Dozio E, Boghossian S, Bovo G, Martos Riaño V, et al. Activation of the Y1 receptor by neuropeptide Y regulates the growth of prostate cancer cells. Endocrinology 2006; 147: 1466–73.
- Magni P, Motta M. Expression of neuropeptide Y receptors in human prostate cancer cells. Ann Oncol 2001; 12 Suppl 2: S27–9.
- Torrealba N, Rodríguez-Berriguete G, Fraile B, Olmedilla G, Martínez-Onsurbe P, et al. Expression of several cytokines in prostate cancer: Correlation with clinical variables of patients. Relationship with biochemical progression of the malignance. Cytokine 2017; 89: 105–15.
- Laberge RM, Sun Y, Orjalo AV, Patil CK, Freund A, et al. MTOR regulates the pro-tumorigenic senescence-associated secretory phenotype by promoting IL1A translation. Nat Cell Biol 2015; 17: 1049–61.
- Kobylewski SE, Henderson KA, Eckhert CD. Identification of ryanodine receptor isoforms in prostate DU-145, LNCaP, and PWR-1E cells. Biochem Biophys Res Commun 2012; 425: 431–5.
- Mariot P, Prevarskaya N, Roudbaraki MM, Le Bourhis X, Van Coppenolle F, et al. Evidence of functional ryanodine receptor involved in apoptosis of prostate cancer (LNCaP) cells. Prostate 2000; 43: 205–14.
- Bettin A, Reyes I, Reyes N. Gene expression profiling of prostate cancer-associated genes identifies fibromodulin as potential novel biomarker for prostate cancer. Int J Biol Markers 2016; 31: e153–62.
- Ashour N, Angulo JC, Andrés G, Alelú R, González-Corpas A, et al. A DNA hypermethylation profile reveals new potential biomarkers for prostate cancer diagnosis and prognosis. Prostate 2014; 74: 1171–82.
- Goo YA, Goodlett DR, Pascal LE, Worthington KD, Vessella RL, et al. Stromal mesenchyme cell genes of the human prostate and bladder. BMC Urol 2005; 5: 17.
- Pio R, Jia Z, Baron VT, Mercola D; UCI NCI SPECS Consortium of the Strategic Partners for the Evaluation of Cancer Signatures-Prostate Cancer. Early growth response 3 (Egr3) is highly over-expressed in non-relapsing prostate cancer but not in relapsing prostate cancer. PLoS One 2013; 8: e54096.
- Floyd BJ, Wilkerson EM, Veling MT, Minogue CE, Xia C, et al. Mitochondrial protein interaction mapping identifies regulators of respiratory chain function. Mol Cell 2016; 63: 621–32.
- Fang TT, Sun XJ, Chen J, Zhao Y, Sun RX, et al. Long non-coding RNAs are differentially expressed in hepatocellular carcinoma cell lines with differing metastatic potential. Asian Pac J Cancer Prev 2014; 15: 10513–24.
- Su A, Ra S, Li X, Zhou J, Binder S. Differentiating cutaneous squamous cell carcinoma and pseudoepitheliomatous hyperplasia by multiplex qRT-PCR. Mod Pathol 2013; 26: 1433–7.
| References|| |
Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin
2020; 70: 7–30.
Chen W, Zheng R, Baade PD, Zhang S, Zeng H, et al.
Cancer statistics in China, 2015. CA Cancer J Clin
2016; 66: 115–32.
Hanahan D, Weinberg RA. The hallmarks of cancer. Cell
2000; 100: 57–70.
Nelson WG, Yegnasubramanian S, Agoston AT, Bastian PJ, Lee BH, et al.
Abnormal DNA methylation, epigenetics, and prostate cancer. Front Biosci
2007; 12: 4254–66.
Aryee MJ, Liu W, Engelmann JC, Nuhn P, Gurel M, et al.
DNA methylation alterations exhibit intraindividual stability and interindividual heterogeneity in prostate cancer metastases. Sci Transl Med
2013; 5: 169ra10.
Yang M, Park JY. DNA methylation in promoter region as biomarkers in prostate cancer. Methods Mol Biol
2012; 863: 67–109.
Li LC, Okino ST, Dahiya R. DNA methylation in prostate cancer. Biochim Biophys Acta
2004; 1704: 87–102.
Formosa A, Lena AM, Markert EK, Cortelli S, Miano R, et al.
DNA methylation silences miR-132 in prostate cancer. Oncogene
2013; 32: 127–34.
Albany C, Alva AS, Aparicio AM, Singal R, Yellapragada S, et al.
Epigenetics in prostate cancer. Prostate Cancer
2011; 2011: 580318.
Truong M, Yang B, Wagner J, Desotelle J, Jarrard DF. Analysis of promoter non-CG methylation in prostate cancer. Epigenomics
2013; 5: 65–71.
Feinberg AP, Tycko B. The history of cancer epigenetics. Nat Rev Cancer
2004; 4: 143–53.
Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet
2002; 3: 415–28.
Mah WC, Lee CG. DNA methylation: potential biomarker in hepatocellular carcinoma. Biomark Res
2014; 2: 5.
Wei SH, Balch C, Paik HH, Kim YS, Baldwin RL, et al.
Prognostic DNA methylation biomarkers in ovarian cancer. Clin Cancer Res
2006; 12: 2788–94.
Liao Y, Xu K. Epigenetic regulation of prostate cancer: the theories and the clinical implications. Asian J Androl
2019; 21: 279–90.
Rodriguez-Paredes M, Esteller M. Cancer epigenetics reaches mainstream oncology. Nat Med
2011; 17: 330–9.
Wang Y, Jadhav RR, Liu J, Wilson D, Chen Y, et al.
Roles of distal and genic methylation in the development of prostate tumorigenesis revealed by genome-wide DNA methylation analysis. Sci Rep
2016; 6: 22051.
Gilad Y, Rifkin SA, Pritchard JK. Revealing the architecture of gene regulation: the promise of eQTL studies. Trends Genet
2008; 24: 408–15.
Bryois J, Buil A, Evans DM, Kemp JP, Montgomery SB, et al
. Cis and trans effects of human genomic variants on gene expression. PLoS Genet
2014; 10: e1004461.
Westra H, Peters MJ, Esko T, Yaghootkar H, Schurmann C, et al
. Systematic identification of trans eQTLs as putative drivers of known disease associations. Nat Genet
2013; 45: 1238–43.
Huan T, Joehanes R, Song C, Peng F, Guo Y, et al
. Genome-wide identification of DNA methylation QTLs in whole blood highlights pathways for cardiovascular disease. Nat Commun
2019; 10: 4267.
Gupta R, van Dongen J, Fu Y, Abdellaoui A, Tyndale RF, et al
. Epigenome-wide association study of serum cotinine in current smokers reveals novel genetically driven loci. Clin Epigenetics
2019; 11: 1.
Gong J, Wan H, Mei S, Ruan H, Zhang Z, et al
. Pancan-meQTL: a database to systematically evaluate the effects of genetic variants on methylation in human cancer. Nucleic Acids Res
2019; 47: D1066–72.
Li Q, Seo JH, Stranger B, McKenna A, Pe'Er I, et al.
Integrative eQTL-based analyses reveal the biology of breast cancer risk loci. Cell
2013; 152: 633–41.
Thibodeau SN, French AJ, McDonnell SK, Cheville J, Middha S, et al.
Identification of candidate genes for prostate cancer-risk SNPs utilizing a normal prostate tissue eQTL data set. Nat Commun
2015; 6: 8653.
Che J, Shin M. A meta-analysis strategy for gene prioritization using gene expression, SNP genotype, and eQTL data. Biomed Res Int
2015; 2015: 576349.
Li Q, Stram A, Chen C, Kar S, Gayther S, et al.
Expression QTL-based analyses reveal candidate causal genes and loci across five tumor types. Hum Mol Genet
2014; 23: 5294–302.
Jiang J, Cui W, Vongsangnak W, Hu G, Shen B. Post genome-wide association studies functional characterization of prostate cancer risk loci. BMC Genomics
2013; 14 Suppl 8: S9.
Orozco LD, Cokus SJ, Ghazalpour A, Ingram-Drake L, Wang S, et al.
Copy number variation influences gene expression and metabolic traits in mice. Hum Mol Genet
2009; 18: 4118–29.
Wagner JR, Busche S, Ge B, Kwan T, Pastinen T, et al.
The relationship between DNA methylation, genetic and expression inter-individual variation in untransformed human fibroblasts. Genome Biol
2014; 15: R37.
Yang IV, Pedersen BS, Rabinovich E, Hennessy CE, Davidson EJ, et al.
Relationship of DNA methylation and gene expression in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med
2014; 190: 1263–72.
Kendziorski C, Wang P. A review of statistical methods for expression quantitative trait loci mapping. Mamm Genome
2006; 17: 509–17.
Xu J, Mo Z, Ye D, Wang M, Liu F, et al.
Genome-wide association study in Chinese men identifies two new prostate cancer risk loci at 9q31.2 and 19q13.4. Nat Genet
2012; 44: 1231–5.
Ishkanian AS, Mallof CA, Ho J, Meng A, Albert M, et al.
High-resolution array CGH identifies novel regions of genomic alteration in intermediate-risk prostate cancer. Prostate
2009; 69: 1091–100.
Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc
2009; 4: 44–57.
Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, et al.
Gene ontology: tool for the unification of biology. Nat Genet
2000; 25: 25–9.
Majumdar S, Buckles E, Estrada J, Koochekpour S. Aberrant DNA methylation and prostate cancer. Curr Genomics
2011; 12: 486–505.
Jeronimo C, Henrique R. Epigenetic biomarkers in urological tumors: a systematic review. Cancer Lett
2014; 342: 264–74.
Jeronimo C, Bastian PJ, Bjartell A, Carbone GM, Catto JW, et al.
Epigenetics in prostate cancer: biologic and clinical relevance. Eur Urol
2011; 60: 753–66.
Ngollo M, Dagdemir A, Karsli-Ceppioglu S, Judes G, Pajon A, et al.
Epigenetic modifications in prostate cancer. Epigenomics
2014; 6: 415–26.
Wu P, Cao Z, Wu S. New progress of epigenetic biomarkers in urological cancer. Dis Markers
2016; 2016: 9864047.
Cucchiara V, Yang JC, Mirone V, Gao AC, Rosenfeld MG, et al.
Epigenomic regulation of androgen receptor signaling: potential role in prostate cancer therapy. Cancers (Basel)
2017; 9: 9.
Vaidyanathan G, Cismowski MJ, Wang G, Vincent TS, Brown KD, et al.
The Ras-related protein AGS1/RASD1 suppresses cell growth. Oncogene
2004; 23: 5858–63.
Hannon E, Spiers H, Viana J, Pidsley R, Burrage J, et al.
Methylation QTLs in the developing brain and their enrichment in schizophrenia risk loci. Nat Neurosci
2016; 19: 48–54.
Kato N, Loh M, Takeuchi F, Verweij N, Wang X, et al.
Trans-ancestry genome-wide association study identifies 12 genetic loci influencing blood pressure and implicates a role for DNA methylation. Nat Genet
2015; 47: 1282–93.
Keil KP, Vezina CM. DNA methylation as a dynamic regulator of development and disease processes: spotlight on the prostate. Epigenomics
2015; 7: 413–25.
Geybels MS, Zhao S, Wong CJ, Bibikova M, Klotzle B, et al.
Epigenomic profiling of DNA methylation in paired prostate cancer versus adjacent benign tissue. Prostate
2015; 75: 1941–50.
Barry KH, Moore LE, Sampson JN, Koutros S, Yan L, et al
. Prospective study of DNA methylation at chromosome 8q24 in peripheral blood and prostate cancer risk. Br J Cancer
2017; 116: 1470–9.
FitzGerald LM, Naeem H, Makalic E, Schmidt DF, Dowty JG, et al
. Genome-wide measures of peripheral blood DNA methylation and prostate cancer risk in a prospective nested case-control study. Prostate
2017; 77: 471–8.
Tam L, McGlynn LM, Traynor P, Mukherjee R, Bartlett JM, et al.
Expression levels of the JAK/STAT pathway in the transition from hormone-sensitive to hormone-refractory prostate cancer. Br J Cancer
2007; 97: 378–83.
Amankwah EK, Sellers TA, Park JY. Gene variants in the angiogenesis pathway and prostate cancer. Carcinogenesis
2012; 33: 1259–69.
Hsu FC, Sun J, Wiklund F, Isaacs SD, Wiley KE, et al.
A novel prostate cancer susceptibility locus at 19q13. Cancer Res
2009; 69: 2720–3.
Dong JT. Chromosomal deletions and tumor suppressor genes in prostate cancer. Cancer Metastasis Rev
2001; 20: 173–93.
Kai K, Zhang Z, Yamashita H, Yamamoto Y, Miura Y, et al.
Loss of heterozygosity at the ATBF1-A
locus located in the 16q22 minimal region in breast cancer. BMC Cancer
2008; 8: 262.
Heyn H, Sayols S, Moutinho C, Vidal E, Sanchez-Mut JV, et al.
Linkage of DNA methylation quantitative trait loci to human cancer risk. Cell Rep
2014; 7: 331–8.
Amangyeld T, Shin YK, Lee M, Kwon B, Seo YS. Human MUS81-EME2 can cleave a variety of DNA structures including intact Holliday junction and nicked duplex. Nucleic Acids Res
2014; 42: 5846–62.
Golzio C, Willer J, Talkowski ME, Oh EC, Taniguchi Y, et al.
KCTD13 is a major driver of mirrored neuroanatomical phenotypes of the 16p11.2 copy number variant. Nature
2012; 485: 363–7.
Titus AJ, Way GP, Johnson KC, Christensen BC. Deconvolution of DNA methylation identifies differentially methylated gene regions on 1p36 across breast cancer subtypes. Sci Rep
2017; 7: 11594.
Li D, Roberts R. WD-repeat proteins: structure characteristics, biological function, and their involvement in human diseases. Cell Mol Life Sci
2001; 58: 2085–97.
Wei JL, Fu ZX, Fang M, Zhou QY, Zhao QN, et al.
High expression of CASK correlates with progression and poor prognosis of colorectal cancer. Tumour Biol
2014; 35: 9185–94.
Gray PJ, Lin CC, Cooperberg MR, Jemal A, Efstathiou JA. Temporal trends and the impact of race, insurance, and socioeconomic status in the management of localized prostate cancer. Eur Urol
2017; 71: 729–37.
Lamb AD, Bryant RJ, Camilleri P, Hamdy FC. Orient expression: solving the mystery of Asian prostate cancer? Eur Urol
2018; 73: 340–2.
Shim JS, Park HM, Lee J, Kwon HJ. Global and focused transcriptional profiling of small molecule aminopeptidase N inhibitor reveals its mechanism of angiogenesis inhibition. Biochem Biophys Res Commun
2008; 371: 99–103.
Dall'Era MA, True LD, Siegel AF, Porter MP, Sherertz TM, et al.
Differential expression of CD10 in prostate cancer and its clinical implication. BMC Urol
2007; 7: 3.
Yan C, Boyd DD. ATF3 regulates the stability of p53: a link to cancer. Cell Cycle
2006; 5: 926–9.
Pelzer AE, Bektic J, Haag P, Berger AP, Pycha A, et al.
The expression of transcription factor activating transcription factor 3 in the human prostate and its regulation by androgen in prostate cancer. J Urol
2006; 175: 1517–22.
Krajewska M, Krajewski S, Epstein JI, Shabaik A, Sauvageot J, et al.
Immunohistochemical analysis of bcl
-2, bax, bcl
-X, and mcl
-1 expression in prostate cancers. Am J Pathol
1996; 148: 1567–76.
O'Donovan KJ, Tourtellotte WG, Millbrandt J, Baraban JM. The EGR family of transcription-regulatory factors: progress at the interface of molecular and systems neuroscience. Trends Neurosci
1999; 22: 167–73.
O'Neill D, Jones D, Wade M, Grey J, Nakjang S, et al.
Development and exploitation of a novel mutant androgen receptor modelling strategy to identify new targets for advanced prostate cancer therapy. Oncotarget
2015; 6: 26029–40.
Liu XJ, Li YQ, Chen QY, Xiao SJ, Zeng SE. Up-regulating of RASD1 and apoptosis of DU-145 human prostate cancer cells induced by formononetin in vitro
. Asian Pac J Cancer Prev
2014; 15: 2835–9.
Febbo PG, Mulligan MG, Slonina DA, Stegmaier K, Di Vizio D, et al.
Literature Lab: a method of automated literature interrogation to infer biology from microarray analysis. BMC Genomics
2007; 8: 461.
Shijubo N, Kawabata I, Sato N, Itoh Y. Clinical aspects of Clara cell 10-kDa protein/ uteroglobin (secretoglobin 1A1). Curr Pharm Des
2003; 9: 1139–49.
Bilen MA, Reyes A, Bhowmick D, Maa A, Bast RJ, et al.
Variant prostate carcinoma and elevated serum CA-125. Can J Urol
2014; 21: 7442–8.
Hessels D, Schalken JA. Urinary biomarkers for prostate cancer: a review. Asian J Androl
2013; 15: 333–9.
Song YH, Shiota M, Kuroiwa K, Naito S, Oda Y. The important role of glycine N-methyltransferase in the carcinogenesis and progression of prostate cancer. Mod Pathol
2011; 24: 1272–80.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]