Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Integrated Analysis Reveals together miR-182, miR-200c and miR-221 Can Help in the Diagnosis of Prostate Cancer

  • Yinmin Gu ,

    Contributed equally to this work with: Yinmin Gu, Danqing Lei

    Affiliation Experimental Center of Medical Sciences, Guangxi Medical University, Nanning, Guangxi, China

  • Danqing Lei ,

    Contributed equally to this work with: Yinmin Gu, Danqing Lei

    Affiliation Experimental Center of Medical Sciences, Guangxi Medical University, Nanning, Guangxi, China

  • Xia Qin,

    Affiliation Dermatological Department, First Affiliated Hospital of Guangxi Medical University, Nanning, Guangxi, China

  • Panyu Chen,

    Affiliation Experimental Center of Medical Sciences, Guangxi Medical University, Nanning, Guangxi, China

  • Yi ming Zou ,

    ymzou@uwm.edu (YMZ); ylhupost@163.com (YLH)

    Affiliation Department of Mathematical Sciences, University of Wisconsin-Milwaukee, Milwaukee, Wisconsin, United States of America

  • Yanling Hu

    ymzou@uwm.edu (YMZ); ylhupost@163.com (YLH)

    Affiliations Experimental Center of Medical Sciences, Guangxi Medical University, Nanning, Guangxi, China, Center for Genomic and Personalized Medicine, Guangxi Medical University, Nanning, Guangxi, China

Integrated Analysis Reveals together miR-182, miR-200c and miR-221 Can Help in the Diagnosis of Prostate Cancer

  • Yinmin Gu, 
  • Danqing Lei, 
  • Xia Qin, 
  • Panyu Chen, 
  • Yi ming Zou, 
  • Yanling Hu
PLOS
x

Abstract

Research has shown that microRNAs are promising biomarkers that can be used to promote a more accurate diagnosis of cancer. In this study, we developed an integrated multi-step selection process to analyze available high-throughput datasets to obtain information on microRNAs as cancer biomarkers. Applying this approach to the microRNA expression profiles of prostate cancer and the datasets in The Cancer Genome Atlas Data Portal, we identified miRNA-182, miRNA-200c and miRNA-221 as possible biomarkers for prostate cancer. The associations between the expressions of these three microRNAs with clinical parameters as well as their diagnostic capability were studied. Several online databases were used to predict the target genes of these three microRNAs, and the results were confirmed by significant statistical correlations. Comparing with the other 18 types of cancers listed in The Cancer Genome Atlas Data Portal, we found that the combination of both miRNA-182 and miRNA-200c being up-regulated and miRNA-221 being down-regulated only happens in prostate cancer. This provides a unique biological characteristic for prostate cancer that can potentially be used for diagnosis based on tissue testing. In addition, our study also revealed that these three microRNAs are associated with the pathological status of prostate cancer.

Introduction

Prostate cancer (PCa) is the second most frequently diagnosed cancer and is the sixth highest cause of cancer-related death among men worldwide [1]. It is a clinically heterogeneous-multifocal disease, and the number of cases is steadily increasing [2]. So far, prostate-specific antigen (PSA) detection has provided the most effective biomarker for diagnosis and the response to treatment in PCa. However, the sensitivity and specificity of PSA testing are insufficient, which results in low detection rates [3]. With the advancement of research on carcinogenesis, PCa studies have increasingly focused on new strategies for early detection and prevention [4].

Studies have suggested that microRNA (miRNA), a type of endogenous, small, non-coding RNA with an approximate length of 22 nucleotides [5], may be linked to cancer; specifically, aberrant miRNAs are linked to clinical behavior and they can be promising biomarkers for more accurate diagnostic/prognosis of cancers [6,7,8]. Using miRNAs as potential diagnostic markers in PCa has been reported in the literature. It has been reported that miR-141 is elevated in the serum of PCa patients and correlates significantly with PSA [9]. Moreover, it has been shown that a five-miRNA panel (downregulation of let-7e, let-7c and miR-30c, upregulation of miR-622 and miR-1285) is capable of accurately differentiating PCa from benign prostate hyperplasia (BPH) and normal samples [10]. These reports suggested that identifying aberrations in miRNAs associated with a particular type of cancer should provide good biomarkers for these specific cancers and promote earlier diagnosis.

Today, high-throughput technologies have produced a large amount of cancer data, so it is desirable to use these data to identify miRNAs and the aberrations that are associated with different types of cancers. However, the analyses of these high-throughput data face difficulties. One difficulty is the lack of homogeneity among different sets of miRNA data due to the fact that different platforms were used to acquire them. Different sets of miRNA data that are expressed differently tend to show inconsistencies with each other. Among the methods developed to address this problem, the robust rank aggregation (RRA) method, which defines the rank vector for each gene based only on the datasets where it is present, has been shown to provide statistically significant miRNA meta-signatures [11, 12]. The method is based on using order statistics to compare each gene to the baseline case, where all the preference lists are randomly shuffled, and then assigns significance levels to the findings [11]. However, in order to accurately identify the potentially useful miRNAs as biomarkers from the selected miRNAs using RRA, further statistical analysis and verification are necessary in addition to the RRA method.

In this study, we applied a new multi-step selection approach to the existing high-throughput data of PCa for the purpose of identifying miRNAs as PCa biomarkers. We first applied the RRA method to select potential miRNA biomarkers in prostate tumors using 11 published miRNA expression profiles. Then, we used The Cancer Genome Data Atlas (TCGA) to further verify the selected miRNAs in multiple ways by the Wilcoxon rank sum test. We found that the combination of two up-regulated miRNAs (miRNA-182, miRNA-200c) and one down-regulated miRNA (miRNA-221) is unique in PCa. This suggests that this combination of miRNAs and their expression levels could potentially provide additional effective diagnostic indicators for PCa.

Materials and Methods

Literature search

The information on miRNA expression profiling studies on PCa was systematically searched in PubMed, Embase and Highwire databases, using the search string (prostate and (cancer* OR tumor* OR tumour*) and (mirna* OR microrna * OR mir-*)). In addition, we obtained miRNA expression profiles for PCa through searching the Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo/) [13] and ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) repositories [14]. The search was restricted to data published between January 1, 2005 and January 31, 2014. Our selection criteria were: (a) original experimental articles providing a comparison of prostate tumor tissues and non-tumor tissues, (b) studies were about miRNA expressions, (c) the studied organism was Homo sapiens and (d) viral miRNAs and non-miRNA probes were excluded.

In this study, the non-tumor tissues included prostate tissues adjacent to a tumor and tissues from independent healthy donors, but did not include BPH. The data extracted from each study included: first author, Gleason score, region, assay type, the number of miRNA probes and the number of samples. The lists of miRNAs with statistically significant expressions were either extracted from the publications or obtained from the authors directly.

Accessing and processing of TCGA data

The data including normalized miRNA-HiSeq expression values, raw read counts of mRNA-seq and clinical information for PCa were downloaded by the “TCGA-Assembler” package in R [15]. Level 3 HiSeq genes normalized data for PCa were acquired from FIREHOSE Broad GDAC (http://gdac.broadinstitute.org/). Reads per kilobase of exon model per million mapped (RPKM) normalized values for mRNA expressions and reads per million miRNA mapped (RPM) normalized values for miRNA expressions were further log2-transformed. Fold changes in miRNA expressions between tumors and normal tissues were calculated using median-centered RPM values. Based on raw read counts, differential mRNA expression analysis was performed by the DESeq Bioconductor package in R [16], which uses a negative binomial distribution model and local regression to estimate the relationship between the mean and variance of each gene. All differentially expressed genes were considered significant if the absolute values of their log2 fold changes were greater than 1 and the false discovery rates (FDR) were < 0.1.

Prediction of the target genes of miRNA and enrichment analysis

Target gene predictions of differentially expressed miRNAs were performed using TargetScan [17], miRwalk [18] and the PICTAR database [19]. The predicted targets must have been selected by at least two algorithms. Validated targets from the CLIP-Seq database starBase [20] were also used in our selection process. The selected target genes were the overlapping targets of the predicted targets and the validated target genes. The target genes in the following discussions show significant correlations with the expressions of miRNAs. The DAVID tool [21] was used to elucidate the molecular functions of the candidate miRNAs.

Statistical analysis

The RRA method we used was adopted from a previous study [12]. We integrated 11 miRNA lists to make sure that the lists were ranked consistently better than expected by the RobustRankAggreg package in R [11] using the function “aggregateRanks”. The extracted miRNA lists were first prioritized based on statistical test p-values (less than 0.05 was considered significant). If the p-value was not reported, we used the fold change (FC) instead. Leave-one-out cross-validation (LOOCV) was performed after the RRA analysis to assess the stability of the acquired p-values. We found that the p-values stabilized after 10,000 runs of this analysis. We thus used 10,000 tests as the cutoff, excluded one random miRNA list for each test, and used the average p-value of each miRNA as the final p-value.

Then, the Spearman’s rank correlation coefficients and the two-tailed p-values of the miRNAs were estimated using the “cor.test” function in R. The relationships among clinical features and the expressions of the miRNAs were evaluated by Wilcoxon rank sum or Kruskal-Wallis non-parametric test using the “wilcox.test” function or the “kruskal.test” function in R, respectively. The diagnostic capabilities of miRNAs were assessed using the “pROC” package in R. All statistical calculations were performed on log2 transformed expression levels.

Results

Selection of PCa meta-signature miRNA candidates

Our selection criteria (see Materials and Methods) resulted in the selection of 11 miRNA profiling datasets for our analysis (Fig 1). Six of the studies also provided Gleason scores or carcinoma tissues, so we classified them as the carcinoma group. A brief overview of the 11 selected datasets is presented in Table 1. In total, we analyzed 347 carcinoma samples and 188 normal samples. These datasets included various microarray platforms and the number of miRNA probes ranged from 88 to 847. The results from RRA provided eight statistically significant miRNAs consisting of four up-regulated and four down-regulated markers. After the stability of the p-value was tested, miR-375 (p = 0.053 > 0.050) and miR-25-3p (p = 0.074 > 0.050) were excluded, and the other six meta-signature miRNAs (p < 0.050) were kept for further analysis. These six selected miRNAs included two up-regulated miRNAs (miR-182-5p, miR-200c-3p) and four down-regulated miRNAs (miR-145-5p, miR-205-5p, miR-221-3p, and miR-222-3p).

Validation of meta-signature miRNAs using TCGA database

TCGA database was used to validate the expression level changes of the six selected meta-signature miRNAs in PCa patients. In the validation, the corrected p -value cut-off was set to 0.05 and the FC cut-off was set to 2 in order to provide more rigorous identification indicators to differentiate miRNAs. Based on the analysis of 498 PCa samples and 52 normal samples, we found that three statistically significant miRNAs were in agreement with the analysis result of the expression profiling by RRA (Table 2 and Fig 2). Specifically, miR-182 (FDR = 3.50E-26, FC = 5.89) and miR-200c (FDR = 1.10E-26, FC = 3.50) were significantly up-regulated, miR-221 (FDR = 1.43E-16, FC = 0.41) was significantly down-regulated, and the FC of miR-145, miR-205 and miR-222 were greater than 0.5 (FDR = 3.29E-01, 6.50E-06 and 1.33E-12, respectively; FC = 0.90, 0.52 and 0.52, respectively). To further verify our selection of these three miRNAs as potential biomarkers for PCa, we then investigated their expression levels in 18 other tumor types in TCGA data. The analytical results from these three miRNAs in 18 other tumor types and PCa are shown in S1 Table and Fig 3, respectively. These results showed that the event of up-expression of miR-182 and miR-200c and down-expression of miR-221 can only be observed in PCa.

thumbnail
Fig 2. TCGA data for the three differentially expressed miRNAs in prostate primary solid tumor (n = 498) and normal solid tissue (n = 52).

(A) The expression levels of miR-182 in prostate primary solid tumor and normal solid tissue. (B) The expression levels of miR-200c in prostate primary solid tumor and normal solid tissue. (C) The expression levels of miR-145 in prostate primary solid tumor and normal solid tissue. The expression level values were calculated using log2 transformed RPM values.

https://doi.org/10.1371/journal.pone.0140862.g002

thumbnail
Fig 3. The log2 FC values of the three miRNAs in 19 cancer types.

Fold change is the ratio of the median signals between cancer tissue and normal tissue. Abbreviations: FC, Fold change; PRAD, Prostate adenocarcinoma; BLCA, Bladder urothelial carcinoma; CHOL, Cholangiocarcinoma; COAD, Colon adenocarcinoma; ESCA, Esophageal carcinoma; HNSC, Head and neck squamous cell carcinoma; KICH, Kidney chromophobe; KIRC, Kidney renal clear cell carcinoma; KIRP, Kidney renal papillary cell carcinoma; LIHC, Liver hepatocellular carcinoma; LUAD, Lung adenocarcinoma; LUSC, Lung squamous cell carcinoma; PCPG, Pheochromocytoma and paraganglioma; READ, Rectal adenocarcinoma; STAD, Stomach adenocarcinoma; THCA, Thyroid carcinoma; THYM, Thymoma; PAAD, Pancreatic adenocarcinoma; and SKCM, Skin cutaneous melanoma. * The indicated miRNA expressed differentially in a cancer.

https://doi.org/10.1371/journal.pone.0140862.g003

Next, we used the Spearman rank correlation to test the association among the three identified miRNAs in PCa. We found that miR-182 and miR-200c showed a positive correlation (rs = 0.633, p < 2.200E-16), and they were negatively correlated with down-regulated miR-221 (miR-182 vs miR-221: rs = -0.479, p < 2.200E-16; miR-200 vs miR-221: rs = -0.335, p = 1.983E-14). Therefore, we carried out further analysis to confirm these three miRNAs and their unique combination of expression levels in PCa.

To explore the expression levels of miR-182, miR-200c and miR-221 in PCa blood or serum, we used GEO2R [33] to analyze the data, we found that miR-182 (log2 FC = 1.740, p = 0.046) and miR-200c (log2 FC = 1.963, p = 0.009) showed overexpressions and miR-221 (log2 FC = -0.810, p = 0.110) showed no significant difference in the series GSE24201 [34], which included 14 blood samples of PCa patients and 15 samples of healthy brothers from 11 families. We also compared the serums of 3 TRansgenic Adenocarcinoma of Mouse Prostate (TRAMP) form mice to their 3 wild-type littermates from GSE29314 [35], and we found that miR-182 (log2 FC = 1.187, p = 0.010) and miR-200c (log2 FC = 1.389, p = 0.002) were also up-regulated and reached p < 0.05, while miR-221 (log2 FC = -0.047, p = 0.818) did not show a different trend.

Correlation of the expressions of the three selected miRNAs with clinicopathologic parameters

The clinicopathologic features of the PCa patients obtained from TCGA are shown in S2 Table. Their mean age was 60.4±7.0 and their pre-surgery PSA ranged 0.7–87 (lg/l). Our Spearman rank analysis showed that these three miRNAs were significantly associated with pre-surgery PSA levels and ages. Both miR-182 and miR-200c showed a positive correlation with the two clinical parameters while miR-221 showed a negative correlation. These three miRNAs did not show significant differences among races (Table 3). We found that miR-221 showed significant differentiation in terms of Gleason grade (low Gleason score vs. high Gleason score), lymph node positivity, the pathology of the N stage (pN0 vs. pN1), and the pathology of the T stage (pT2 vs. pT3 vs. pT4) (p < 0.05 for all comparisons) (see Table 3), and it was down-regulated in patients with higher Gleason scores, advanced stage tumors and positive lymph nodes. However, we did not find significant differences between these clinical features and miR-182 or miR-200c.

thumbnail
Table 3. Correlations between the expressions of the three miRNAs and clinical features.

https://doi.org/10.1371/journal.pone.0140862.t003

Diagnostic value of the selected miRNAs

We conducted response operating characteristic (ROC) analyses to evaluate the use of the three selected miRNAs as potential biomarkers to differentiate PCa tissues from normal tissues (Fig 4), and we found that MiR-200c (AUC = 0.963, 95% confidence interval, CI = 0.9463–0.980, p < 0.0001) showed a higher diagnostic capability than miR-182 (AUC = 0.957, 95% confidence interval, CI = 0.9248–0.9896, p < 0.0001) and miR-221 (AUC = 0.857, 95% confidence interval, CI = 0.8022–0.9127, p < 0.0001). When a logistic regression approach was used for the combination of these three miRNAs, the ROC curve revealed a much better diagnostic accuracy than if they were used individually, the results showed an AUC of 0.972 (95% confidence interval, CI = 0.9519–0.992, p < 0.0001). In the analysis, the optimal cutoff value was set at a maximal sum of sensitivity and specificity. The diagnostic sensitivity and specificity of the combination of these three miRNAs were found to be 94.2% and 92.6%, respectively.

thumbnail
Fig 4. The areas under the ROC curves for the three identified miRNAs and their combination that differentiate between PCa samples and normal samples.

(A) The ROC curve for miR-182. (B) The ROC curve for miR-200c. (C) The ROC curve for miR-221. (D) The ROC curve for the combination of the three miRNAs.

https://doi.org/10.1371/journal.pone.0140862.g004

Target genes and biological pathway recognition

First, we identified 800 up-regulated and 1698 down-regulated different genes based on a model using the negative binomial distribution for 374 prostate cancers samples and 52 normal controls from TCGA (S3 Table). Second, we predicted target genes for these three miRNAs, and then chose the overlap genes as pathway participants (129 genes for miR-182, 95 genes for miR-200c, and 55 genes for miR-221) (S4 Table). Third, we performed correlation analysis among these three miRNAs and the mRNA expressions of all overlap genes. The results showed that miR-182 and miR-200c were positively associated with almost all up-regulated overlap mRNAs and negatively associated with almost all down-regulated mRNAs. However, the correlation between miR-221 and the overlap genes showed an inverse relation (Fig 5). A surprising observation from our analysis was that regulating synaptic membrane exocytosis 3 (RIMS3) consistently overrepresented with all three miRNAs.

thumbnail
Fig 5. Clusters of miRNA-target genes by their correlation coefficients.

The heat map representation of the correlation patterns for the miRNA targets. The 267 differentially expressed target mRNAs, including 63 up-regulated targets and 204 down-regulated targets, were used to cluster the miRNAs. Rows of the heat map represent target mRNAs, and columns represent the miRNAs. Red squares indicate negative correlations, purple squares indicate positive correlations and white spots (colorless) indicate that there is no correlation.

https://doi.org/10.1371/journal.pone.0140862.g005

To understand the general biological functions of these three miRNAs, we conducted Protein Analysis Through Evolutionary Relationships (PANTHER) pathway analysis in DAVID via the target genes. The results showed that these three miRNAs did not share the same pathways, but they were frequently related to cell signaling pathways (Fig 6).

thumbnail
Fig 6. Common target genes and biological pathways of the three miRNAs.

The purple oval represents RIMS3; the three blue ovals represent the three miRNAs; black straight lines indicate the target effect of miRNAs on gene; dotted lines indicate the involvement of the miRNAs in PANTHER pathways, and their colors reflect the significances of -log10 transformed p values for the pathways.

https://doi.org/10.1371/journal.pone.0140862.g006

Discussion

In recent years, many studies have been devoted to the discovery of miRNA biomarkers and their biological functions (or molecular mechanism) for PCa. In this study, we applied an integrated multi-step selection approach to analyze the cancer datasets from GEO and TCGA and identified three miRNAs and their unique expression combination pattern that only showed in the PCa datasets.

The notion that miRNAs are implicated in cancer is supported by the evidence that miRNAs are located largely in genomic regions associated with cancer or at fragile sites [36, 37]. For PCa, we can obtain the following information from the literature. It is known that miR-200c is located at 12p13.31, and it was reported [38] that the copy numbers of chromosome region 12p13.31-p12.3 are deleted in PCa. It has also been reported [39] that chromosome region 7q32.2 in which miR-182 sits has a high incidence of heterozygosity and/or microsatellite imbalance alterations in prostate carcinogenesis. Histone methylation results in the silencing of the entire miR-221/miR-222 cluster, as clarified by an analysis of the methylation signature in PCa cell lines [40]. Therefore, just based on the published information on these locations, we may form the hypothesis that miR-182, miR-200c and miR-221/miR-222 are closely related to PCa.

It was also reported that miRNA-182 is over-expressed in PCa and plays an active role in proliferation and invasion in vitro and vivo [41, 42]. It has been reported that miR-182 in PCa tissues and four cell lines (LNCap, PC-3, DU145, and 22Rv1) have higher levels than in BPH tissues and normal prostatic epithelial (RWPE-1) cells [41]. Associated with PCa progression, the over-expression of miR-182 represses the expression of the tumor suppressor gene FOXF2, which decreases PCa cell invasion and migration [42]. In prostate cells, miR-182 induces mesenchymal to epithelial transition features and growth factor-independent growth by repressing SNAI2, which has been demonstrated to be a repressor of proliferation in PCa cells [43, 44]. Our target gene prediction study correctly identified FOXF2 and SNAI2. The miRNA-200 family, consisting of five members (miR-200a, miR-200b, miR-200c, miR-429, and miR-141), is treated as a tumor suppressor as it inhibits epithelial-to-mesenchymal transition, tumor cell invasion and metastasis [45]. The miR-200c~141 cluster depresses the proliferation of human metastatic prostate cancer cells by inhibiting JAGGED1, which may be important for metastases [46]. It has been reported [47] that miR-221 was down-regulated in aggressive PCa and associated with the Gleason score, and thus indicated the tumor stage. Down-regulation of miR-221 has also been detected in prostate secretion samples [48]. In addition, miR-221s are thought to be involved in the development or stability of the castration-resistant prostate cancer (CRPC) phenotype because over-expression has been observed in CRPC cells [49]. Also, many repeated validations have indicated that miR-221 in PCa cells has the ability to regulate the expression of p27/kip1 and inhibit several cyclin-dependent kinase complexes [50, 51]. All of these are in accordance with our analysis which showed that the three identified microRNA were involved in the development of PCa.

Increasing in age is correlated with PCa risk [52, 53], since levels of miRNAs change with ageing [54], and PCa diagnosis and treatments have been guided by PSA [55]. Our results showed that age and PSA are positively correlated with up-regulated miRNAs (miR-182 and miR-200c), but negatively related to down-regulated miRNA (miR-221). The combination of multi-biomolecules has been proposed to serve as efficient indicators for diagnosis in many studies [56, 57]. Our results showed that the AUC of the combination of these selected three miRNAs was very high for PCa tissues.

Since tumor cells can release miRNAs into the body’s circulation [58], it is desirable to use miRNAs in body fluids for diagnostic purposes if possible. It has been reported that the levels of miRNA in PCa tissue are highly correlated with pre-prostatectomy levels in serum [54]. However, though the combination of increases in miR-182 and miR-200c expressions and the reduction in miR-221 expression is unique (with statistical significance) in PCa blood, further studies will be needed to differentiate these blood markers in PCa from those in LUSC (Lung squamous cell carcinoma) or in BLCA (Bladder urothelial carcinoma) (see Fig 3 and S1 Table).

Through our analyses of target genes and correlations, we found that down-regulated RIMS3 was present simultaneously with the three identified miRNAs. This has not been reported for PCa before. Also some of the pathways associated with PCa in which the three identified miRNAs participate have been reported before. For examples: (1) It has been reported that, together with TMPRSS2-ERG, the PI3-kinase pathway activates prostate oncogenesis [59]; (2) Wnt signaling and its key component β-catenin contribute to prostate tumorigenesis [60]; and (3) the hedgehog signaling pathway is involved in PCa development and in the progression to more aggressive and even therapy-resistant disease states [61]. All of these studies provide additional supporting evidence for our findings.

Conclusion

To the best of our knowledge, this study is the first to identify and analyze combinations of multiple miRNAs as a potential biomarker for PCa based on the analysis of miRNA microarray and TCGA datasets. Our study shows that an integrated analysis method based on RRA and followed by other statistical analyses and verifications can effectively identify combinations of multiple miRNAs as potential cancer biomarkers. Our results suggest that the three selected miRNAs and their uniquely combined expression pattern in PCa could be of clinical value in addition to existing testing methods. Finally, although we only applied the proposed multi-step selection approach to study high-throughput datasets for the purpose of identifying potential biomarkers for PCa, we believe that this approach can also be applied to investigate other cancer types using similar high-throughput datasets.

Supporting Information

S1 Table. The expression levels of miR-182, miR-200c and miR-221 in PCa and other 18 cancer types.

https://doi.org/10.1371/journal.pone.0140862.s002

(XLSX)

S2 Table. The demographic information of the PCa patients in the datasets of TCGA.

https://doi.org/10.1371/journal.pone.0140862.s003

(DOCX)

S3 Table. The 800 up-regulated and the 1698 down-regulated different genes identified in PCa samples and normal samples.

https://doi.org/10.1371/journal.pone.0140862.s004

(XLSX)

S4 Table. The result of using Spearman rank correlation to test the association between expression levels of miRNA-182, miR-200c and miR-221 and their predicted targets in PCa patients (n = 481).

https://doi.org/10.1371/journal.pone.0140862.s005

(XLSX)

Acknowledgments

The main part of this work was done during the visit of the first and the last authors to the University of Wisconsin-Milwaukee. They wish to thank the University of Wisconsin-Milwaukee and its faculty for the hospitality they received during their visit.

Author Contributions

Conceived and designed the experiments: YMG YLH YMZ. Performed the experiments: XQ YMG. Analyzed the data: PYC YMG. Contributed reagents/materials/analysis tools: DQL. Wrote the paper: YMG.

References

  1. 1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011; 61(2): 69–90. pmid:21296855
  2. 2. He HC, Zhu JG, Chen XB, Chen SM, Han ZD, Dai QS, et al. MicroRNA-23b downregulates peroxiredoxin III in human prostate cancer. FEBS Lett 2012; 586(16): 2451–8. pmid:22710126
  3. 3. Sita-Lumsden A, Dart DA, Waxman J, Bevan CL. Circulating microRNAs as potential new biomarkers for prostate cancer. Br J Cancer 2013; 108(10): 1925–30. pmid:23632485
  4. 4. Fan C, Chen C, Wu D. The association between common genetic variant of microRNA-499 and cancer susceptibility: a meta-analysis. Mol Biol Rep 2013; 40(4): 3389–94. pmid:23271127
  5. 5. Zamore PD, Haley B. Ribo-gnome: the big world of small RNAs. Science 2005; 309(5740): 1519–-24. pmid:16141061
  6. 6. Martens-Uzunova ES, Jalava SE, Dits NF, van Leenders GJ, Møller S, Trapman J, et al. Diagnostic and prognostic signatures from the small non-coding RNA transcriptome in prostate cancer. Oncogene 2012; 31(8): 978–91. pmid:21765474
  7. 7. Carlsson J, Helenius G, Karlsson MG, Andrén O, Klinga-Levan K, Olsson B. Differences in microRNA expression during tumor development in the transition and peripheral zones of the prostate. BMC Cancer 2013; 13: 362. pmid:23890084
  8. 8. Walter BA, Valera VA, Pinto PA, Merino MJ. Comprehensive microRNA profiling of prostate cancer. J Cancer 2013; 4(5): 350–7.pmid:23781281
  9. 9. Mitchell PS, Parkin RK, Kroh EM, Fritz BR, Wyman SK, Pogosova-Agadjanyan EL, et al. Circulating microRNAs as stable blood-based markers for cancer detection. Proc Natl Acad Sci U S A 2008; 105(30): 10513–8. pmid:18663219
  10. 10. Chen ZH, Zhang GL, Li HR, Luo JD, Li ZX, Chen GM, et al. A panel of five circulating microRNAs as potential biomarkers for prostate cancer. Prostate 2012; 72(13): 1443–1452. pmid:22298030
  11. 11. Kolde R, Laur S, Adler P, Vilo J. Robust rank aggregation for gene list integration and meta-analysis. Bioinformatics 2012; 28(4): 573–80. pmid:22247279
  12. 12. Võsa U, Vooder T, Kolde R, Vilo J, Metspalu A, Annilo T. Meta-analysis of microRNA expression in lung cancer. Int J Cancer 2013; 132(12): 2884–2893. pmid:23225545
  13. 13. Barrett T, Troup DB, Wilhite SE, Ledoux P, Rudnev D, Evangelista C, et al. NCBI GEO: archive for high-throughput functional genomic data. Nucleic Acids Res 2009; 37 (Database issue): D885–90. pmid:18940857
  14. 14. Parkinson H, Kapushesky M, Kolesnikov N, Rustici G, Shojatalab M, Abeygunawardena N, et al. ArrayExpress updata-from an archive of functional genomics experiments to the atlas of gene expression. Nucleic Acids Res 2009; 37 (Database issue): D868–72. pmid:19015125
  15. 15. Zhu Y, Qiu P, Ji Y. TCGA-assembler: open-source software for retrieving and processing TCGA data. Nat Methods 2014; 11(6): 599–600. pmid:24874569
  16. 16. Anders S, Huber W. Differential expression analysis for sequence count data. Anders and Huber Genome Biology 2010; 11(10): R106. pmid:20979621
  17. 17. Lewis BP, Shih IH, Jones-Rhoades MW, Bartel DP, Burge CB. Prediction of Mammalian MicroRNA Targets. Cell 2003; 115(7): 787. pmid:14697198
  18. 18. Dweep H, Sticht C, Pandey P, Gretz N. miRWalk-database: prediction of possible miRNA binding sites by "walking" the genes of three genomes. J Biomed Inform 2011; 44(5): 839–47. pmid:21605702
  19. 19. Krek A, Grün D, Poy MN, Wolf R, Rosenberg L, Epstein EJ, et al. Combinatorial microRNA target predictions. Nat Genet 2005; 37(5): 495–500. pmid:15806104
  20. 20. Yang JH, Li JH, Shao P, Zhou H, Chen YQ, Qu LH. starBase: a database for exploring microRNA-mRNA interaction maps from Argonaute CLIP-Seq and Degradome-Seq data; 39(Database issue): D202–9. pmid:21037263
  21. 21. Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, et al. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 2003; 4(5): P3. pmid:12734009
  22. 22. Ambs S, Prueitt RL, Yi M, Hudson RS, Howe TM, Petrocca F, et al. Genomic profiling of microRNA and messenger RNA reveals deregulated microRNA expression in prostate cancer cellular regulate signaling and microRNAs. Cancer Res 2008; 68(15): 6162–70.pmid:18676839
  23. 23. Lin PC, Chiu YL, Banerjee S, Park K, Mosquera JM, Giannopoulou E, et al. Epigenetic repression of miR-31 disrupts androgen receptor homeostasis and contributes to prostate cancer progression. Cancer Res 2013; 73 (3): 1232–44.pmid:23233736
  24. 24. Wach S, Nolte E, Szczyrba J, Stöhr R, Hartmann A, Ørntoft T, et al. MicroRNA profiles of prostate carcinoma detected by multiplatform microRNA screening. Int J Cancer 2012; 130(3): 611–21. pmid:21400514
  25. 25. Schaefer A, Jung M, Mollenkopf HJ, Wagner I, Stephan C, Jentzmik F, et al. Diagnostic and prognostic implications of microRNA profiling in prostate carcinoma. Int J Cancer 2010; 126(5): 1166–76. pmid:19676045
  26. 26. Martens-Uzunova ES, Jalava SE, Dits NF, van Leenders GJ, Møller S, Trapman J, et al. Diagnostic and prognostic signatures from the small non-coding RNA transcriptome in prostate cancer. Oncogene 2012; 31(8): 978–91. pmid:21765474
  27. 27. Carlsson J, Davidsson S, Helenius G, Karlsson M, Lubovac Z, Andrén O, et al. A miRNA expression signature that separates between normal and malignant prostate tissues. Cancer Cell Int 2011; 11(1): 14. pmid:21619623
  28. 28. Carlsson J, Helenius G, Karlsson MG, Andrén O, Klinga-Levan K, Olsson B. Differences in microRNA expression during tumor development in the transition and peripheral zones of the prostate. BMC Cancer 2013; 13:362. pmid:23890084
  29. 29. Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, et al. A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci U S A 2006; 103(7): 2257–61. pmid:16461460
  30. 30. Tong AW, Fulgham P, Jay C, Chen P, Khalil I, Liu S, et al. MicroRNA profile analysis of human prostate cancers. Cancer Gene Ther 2009; 16(3): 206–16. pmid:18949015
  31. 31. Ozen M, Creighton CJ, Ozdemir M, Ittmann M. Widespread deregulation of microRNA expression in human prostate cancer. Oncogene 2008; 27(12): 1788–93. pmid:17891175
  32. 32. Walter BA, Valera VA, Pinto PA, Merino MJ. Comprehensive microRNA Profiling of Prostate Cancer. J Cancer 2013; 4: 350–7. pmid:23781281
  33. 33. Barrett T, Wilhite SE, Ledoux P, Evangelista C, Kim IF, Tomashevsky M, et al. NCBI GEO: archive for functional genomics data sets—update. Nucl Acids Res 2013; 41 (Database issue): D991–D995. pmid:23193258
  34. 34. Mattila H, Schindler M, Isotalo J, Ikonen T, Vihinen M, Oja H, et al. NMD and microRNA expression profiling of the HPCX1 locus reveal MAGEC1 as a candidate prostate cancer predisposition gene. BMC Cancer 2011 Aug 2;11:327. pmid:21810217
  35. 35. Selth LA, Townley S, Gillis JL, Ochnik AM, Murti K, Macfarlane RJ, et al. Discovery of circulating microRNAs associated with human prostate cancer using a mouse model of disease. Int J Cancer 2012 Aug 1;131(3):652–61. pmid:22052531
  36. 36. Esquela-Kerscher A, Slack FJ. Oncomirs-microRNAs with a role in cancer. Nat Rev Cancer 2006; 6(4): 259–69. pmid:16557279
  37. 37. Calin GA, Sevignani C, Dumitru CD, Hyslop T, Noch E, Yendamuri S, et al. Human microRNA genes are frequently located at fragile sites and genomic regions involved in cancers. Proc Natl Acad Sci U S A 2004; 101(9): 2999–3004. pmid:14973191
  38. 38. Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS, et al. Integrative genomic profiling of human prostate cancer. Cancer Cel l 2010; 18(1): 11–22. pmid:20579941
  39. 39. Bryś M, Migdalska-Sęk M, Pastuszak-Lewandoska D, Forma E, Czarnecka K, Domańska D, et al. Diagnostic value of DNA alteration: loss of heterozygosity or allelic imbalance-promising for molecular staging of prostate cancers. Med Oncol 2013; 30(1): 391. pmid:23288724
  40. 40. Ke XS, Qu Y, Rostad K, Li WC, Lin B, Halvorsen OJ, et al. Genome-wide profiling of histone h3 lysine 4 and lysine 27 trimethylation reveals an epigenetic signature in prostate carcinogenesis. PLoS One 2009; 4(3): e4687. pmid:19262738
  41. 41. Liu R, Li J, Teng Z, Zhang Z, Xu Y. Overexpressed microRNA-182 promotes proliferation and invasion in prostate cancer PC-3 cells by down-regulating N-myc downstream regulated gene 1 (NDRG1). PLoS One 2013; 8(7): e68982. pmid:23874837
  42. 42. Hirata H, Ueno K, Shahryari V, Deng G, Tanaka Y, Tabatabai ZL, et al. MicroRNA-182-5p promotes cell invasion and proliferation by down regulating FOXF2, RECK and MTSS1 genes in human prostate cancer. PLoS One 2013; 8(1): e5550. pmid:23383207
  43. 43. Qu Y, Li WC, Hellem MR, Rostad K, Popa M, McCormack E, et al. MiR-182 and miR-203 induce mesenchymal to epithelial transition and self-sufficiency of growth signals via repressing SNAI2 in prostate cells. Int J Cancer 2013; 133(3): 544–55. pmid:23354685
  44. 44. Liu J, Uygur B, Zhang Z, Shao L, Romero D, Vary C, et al. Slug inhibits proliferation of human prostate cancer cells via downregulation of cyclin D1 expression. Prostate 2010; 70(16): 1768–77. pmid:20564361
  45. 45. Feng X, Wang Z, Fillmore R, Xi Y. MiR-200, a new star miRNA in human cancer. Cancer Letters 2014; 344(2): 166–173. pmid:24262661
  46. 46. Vallejo DM, Caparros E, Dominguez M. Targeting Notch signaling by the conserved miR-8/200 microRNA family in development and cancer cells. EMBO J 2011; 30(4): 756–69. pmid:21224847
  47. 47. Spahn M, Kneitz S, Scholz CJ, Stenger N, Rüdiger T, Ströbel P, et al. Expression of microRNA-221 is progressively reduced in aggressive prostate cancer and metastasis and predicts clinical recurrence. Int J Cancer 2010; 127(2): 394–403. pmid:19585579
  48. 48. Guzel E, Karatas OF, Semercioz A, Ekici S, Aykan S, Yentur S, et al. Identification of microRNAs differentially expressed in prostatic secretions of patients with prostate cancer. Int J Cancer 2015; 136(4): 875–9. pmid:24976077
  49. 49. Sun T, Wang Q, Balk S, Brown M, Lee GS, Kantoff P. The role of microRNA-221 and microRNA-222 in androgen-independent prostate cancer cell lines. Cancer Res 2009; 69(8): 3356–3363. pmid:19351832
  50. 50. Galardi S, Mercatelli N, Giorda E, Massalini S, Frajese GV, Ciafrè SA, et al. MiR-221 and miR-222 expression affects the proliferation potential of human prostate carcinoma cell lines by targeting p27Kip1. J Biol Chem 2007; 282(32): 23716–24. pmid:17569667
  51. 51. Chu IM, Hengst L, Slingerland JM. The Cdk inhibitor p27 in human cancer: prognostic potential and relevance to anticancer therapy. Nat Rev Cancer 2008; 8(4): 253–67. pmid:18354415
  52. 52. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J Clin 2010; 60(5): 277–300.pmid:20610543
  53. 53. Polednak AP. Trends in prostate carcinoma incidence in Connecticut (1988–1994) by age and race. Cancer 1997; 79(1): 99–103. pmid:8988732
  54. 54. Mahn R, Heukamp LC, Rogenhofer S, von Ruecker A, Muller SC, Ellinger J. Circulating microRNAs (miRNA) in serum of patients with prostate cancer. Urology 2011; 77(5): e9–e16. pmid:21539977
  55. 55. Prensner JR, Rubin MA, Wei JT, Chinnaiyan AM. Beyond PSA: the next generation of prostate cancer biomarkers. Sci Transl Med 2012; 4(127): 127rv3.pmid:22461644
  56. 56. Devaux Y, Vausort M, Goretti E, Nazarov PV, Azuaje F, Gilson G, et al. Use of circulating microRNAs to diagnose acute myocardial infarction. Clin Chem 2012; 58(3): 559–67. pmid:22252325
  57. 57. Schaefer A, Jung M, Mollenkopf HJ, Wagner I, Stephan C, Jentzmik F, et al. Diagnostic and prognostic implications of microRNA profiling in prostate carcinoma. Int J Cancer 2010; 126(5): 1166–76. pmid:19676045
  58. 58. Taylor DD, Gercel-Taylor C. MicroRNA signatures of tumor-derived exosomes as diagnostic biomarkers of ovarian cancer. Gynecol Oncol 2008; 110(1): 13–21. pmid:18589210
  59. 59. King JC, Xu J, Wongvipat J, Hieronymus H, Carver BS, Leung DH, et al. Cooperativity of TMPRSS2-ERG with PI3-kinase pathway activation in prostate oncogenesis. Nat Genet 2009; 41(5): 524–6. pmid:19396167
  60. 60. Verras M, Sun Z.Roles and regulation of Wnt signaling and beta-catenin in prostate cancer. Cancer Letters 2006; 237(1): 22–32. pmid:16023783
  61. 61. Gonnissen A, Isebaert S, Haustermans K. Hedgehog signaling in prostate cancer and its therapeutic implication. Int J Mol Sci 2013; 14(7):13979–4007. pmid:23880852