- Research article
- Open Access
Global gene expression analysis reveals reduced abundance of putative microRNA targets in human prostate tumours
© Sun et al; licensee BioMed Central Ltd. 2009
- Received: 30 July 2008
- Accepted: 26 February 2009
- Published: 26 February 2009
Recently, microRNAs (miRNAs) have taken centre stage in the field of human molecular oncology. Several studies have shown that miRNA profiling analyses offer new possibilities in cancer classification, diagnosis and prognosis. However, the function of miRNAs that are dysregulated in tumours remains largely a mystery. Global analysis of miRNA-target gene expression has helped illuminate the role of miRNAs in developmental gene expression programs, but such an approach has not been reported in cancer transcriptomics.
In this study, we globally analysed the expression patterns of miRNA target genes in prostate cancer by using several public microarray datasets. Intriguingly, we found that, in contrast to global mRNA transcript levels, putative miRNA targets showed a reduced abundance in prostate tumours relative to benign prostate tissue. Additionally, the down-regulation of these miRNA targets positively correlated with the number of types of miRNA target-sites in the 3' untranslated regions of these targets. Further investigation revealed that the globally low expression was mainly driven by the targets of 36 specific miRNAs that were reported to be up-regulated in prostate cancer by a miRNA expression profiling study. We also found that the transcript levels of miRNA targets were lower in androgen-independent prostate cancer than in androgen-dependent prostate cancer. Moreover, when the global analysis was extended to four other cancers, significant differences in transcript levels between miRNA targets and total mRNA backgrounds were found.
Global gene expression analysis, along with further investigation, suggests that miRNA targets have a significantly reduced transcript abundance in prostate cancer, when compared with the combined pool of all mRNAs. The abnormal expression pattern of miRNA targets in human cancer could be a common feature of the human cancer transcriptome. Our study may help to shed new light on the functional roles of miRNAs in cancer transcriptomics.
- Prostate Cancer
- miRNA Target
- Prostate Tumour
- Total mRNAs
- PicTar Target
MicroRNAs are endogenous, approximately 22 nt single-stranded non-coding RNAs that negatively regulate protein expression by binding to the 3' untranslated regions (3' UTR) of messenger RNAs (mRNAs) and inhibiting translation or inducing mRNA degradation or deadenylation . MiRNA genes are expressed as large precursor RNAs, called pri-mRNAs, which may encode multiple miRNAs in a polycistronic arrangement. These precursors are converted into mature miRNAs of 19 to 25 nucleotides by the RNase III enzymes, Drosha (nuclear) and Dicer (cytosolic). MiRNAs have been identified in the genomes of plants, animals and viruses. Human miRNAs have been implicated in a variety of biological processes, and it is estimated that 30% of protein coding genes are regulated by miRNAs .
Recently, the potential role of miRNAs in human cancers has been indicated by several studies, which suggest that aberrations in miRNA expression in cancers may be involved in tumour genesis and progression . Abnormal expression of miRNAs has been found in various cancers, and profiling of miRNAs has been shown to be a more accurate method of classifying cancer subtypes than profiling protein-coding genes. In the case of prostate cancer, a highly prevalent disease in the western world, several miRNA expression profiles have been reported [4–8]. Although some of these studies are not consistent, potentially due to differences in samples or chip platforms, all of them confirmed the widespread dysregulation of miRNAs in prostate cancer. However, the disruption of miRNA expression observed in human cancers needs to be understood by analysing the causes and consequences of the miRNA alterations. Thus far, the causes of miRNA dysregulation are partially known to be the results of three mechanisms: (1) miRNA genes tend to locate in cancer related genomic regions; (2) miRNA expression is epigenetically regulated in cancers; and (3) miRNA processing genes such as Drosha and Dicer are de-regulated in cancers. However, little is known about the consequences of improper regulation of miRNA expression. More work needs to be done to show whether these miRNAs have a direct function in cancer progression or are simply differentially modulated in tumours.
Identifying the genes targeted by miRNA is crucial for understanding the functions of the miRNAs. Based on the conservation of 3' UTRs, which are complementary to the "seed" region (nucleotides 2–7 from the 5' end) of miRNAs, several computational methods have been developed to search for miRNA targets. Some of these methods have been biologically validated and proven to be accurate [2, 9]. According to single miRNA-mRNA target pair information, miRNAs are thought to function as either tumour suppressors or oncogenes. For example, the let-7 family has been shown to target the RAS gene . Down-regulation of the let-7 has been found in lung cancer, and this finding is correlated with a poor prognosis. Thus, reduced expression of let-7 is predicted to promote lung cancer progression. However, the situation is complicated by the fact that miRNAs regulate multiple genes, and a single mRNA can be targeted by several different miRNAs. Therefore, the impact of changes in miRNA expression in cancers is likely to be dependent on the cellular context.
Evidence is emerging that miRNAs can not only repress translation of mRNAs but can also induce their degradation, even if the mRNA target sites have only partial complementarity to the miRNAs. For example, several studies have revealed a correlation between the expression of miRNAs and that of their targets through analysis of mRNA target gene expression profiles and in situ hybridisation [11–13]. A recent publication has reported the impact of miRNAs on global mRNA and protein expression and showed that the regulation of protein-coding genes by miRNAs is quite similar at both the transcript and protein levels . Moreover, Wu et al. have reported that miRNAs increase target mRNA decay rates by promoting rapid deadenylation . Assuming that the miRNAs dysregulated in prostate cancer have an influence on the expression of their targets, analysis of target gene expression may provide clues to the functions of miRNAs in prostate cancer.
In this study, we computationally explored the global expression patterns of miRNA targets in human prostate cancer using several published microarray datasets. Interestingly, in contrast to all the mRNAs with altered expression in prostate cancer relative to benign prostate tissue, the transcript levels of miRNA targets were significantly lower. Closer examination revealed a positive correlation between the reduced abundance of miRNA targets and the number of target-site-types in the 3' UTRs of the target mRNAs. Remarkably, we found that the predicted targets of the up-regulated miRNAs in prostate cancer, reported by a miRNA profiling study, were significantly more likely to be down-regulated in prostate tumours than the predicted targets of all other miRNAs. We also found that the transcript levels of miRNA targets were lower in androgen-independent prostate cancer than in androgen-dependent prostate cancer. Furthermore, the abnormal expression pattern of miRNA targets could be a common feature of the human cancer transcriptome.
Globally reduced transcript levels of miRNA targets relative to total mRNAs in prostate tumours
Gene expression datasets used in this global analysis
Number of samples
AD and AI analysis
17 AD/8 AI
Acute myeloid leukaemia
Given the considerable noise of the gene expression data and miRNA target prediction, we ensured the validity of the above observations using three approaches. Firstly, because microarray results may vary depending on quality of the samples, we collected the gene expression profiles that contained a relatively large number of samples and classified the tissues adjacent to the tumours as benign prostate tissues for all three datasets. Secondly, to rule out the influence of different methods on determining the high, low and unchanged expression groups, two different methods, a and b, were adopted for datasets 1 and 2, where the measurements of matched samples were provided. We also adopted three different cut-off values to determine the mRNAs with altered expression. As shown in Figure 1A, the global low expression of miRNA targets did not vanish when we changed methods or cut-offs, and the RR values actually decreased with the raised cut-offs. For example, in dataset 1, RR = 0.79, 0.76, and 0.71 for low, medium and high cut-offs, respectively (Method a, PicTar targets). Finally, to determine if this observation is robust for different miRNA target predictions, we carried out the same analysis using two different target predictions, which are the leading programs in this field, and obtained similar results. Figure 1C shows that the RR values obtained using the two predictions were highly correlated (Pearson correlation = 0.9691). Additionally, since repression by miRNAs also result in decreased translation, we analysed the protein levels of several miRNA targets using a small proteomic dataset. In this dataset, 64 proteins were reported to be dysregulated in prostate cancer, relative to benign prostate, by employing a high-throughput immunoblot approach . After converting the names of these proteins into RefSeq mRNA IDs (some proteins have more than one RefSeq ID), we mapped all miRNA targets to them. As shown in Figure 1B, the miRNA targets showed a propensity to fall in the decreased protein group (RR < 1), suggesting a reduced protein level of miRNA targets in prostate tumour. Taken together, our data indicate that miRNA target genes are expressed at lower levels in prostate cancer than normal prostate tissue.
Association between transcript abundance of miRNAs and their target mRNAs in prostate cancer
Positive correlation between the reduced abundance of miRNA targets and the number of target-site types in their 3' UTRs in prostate cancer
MiRNA target predictions rely strongly on the sequence characteristics of 3' UTRs, which have known functions in the stability, localisation, and translation of mRNA . Most miRNA targets contain more than one type of target site in their 3' UTRs, implying that stringent regulation by one miRNA is rare. For example, the target mRNAs of the 36 up-regulated miRNAs included a major fraction of all PicTar targets (5514 of 8177) which were shared by 168 miRNAs. Interestingly, we found that the average number of target-site types in the target mRNAs of up-regulated miRNAs (≈ 10) was significantly larger than the average number found in the target mRNAs of all other miRNAs (≈ 2, P < 10-100, Wilcoxon-Man-Whitney test).
We found that the down-regulation of miRNA targets in tumour extracts correlated with the length of the 3' UTRs of these mRNAs (Figure 5B). It is not surprising to find more sites just by chance if the 3' UTR is longer. To ask whether this positive correlation is a side effect of the variations of 3' UTR length, we divided the PicTar targets into eight groups according to the length of their 3' UTRs and calculated the average number of target-site types for each group. Since genes with more miRNA sites would have not only relatively longer 3' UTRs but also significantly higher site densities (sites/kb) of 3' UTR sequence , the longer 3' UTRs does not always contain more target sites. This was seen in Figure 5B, where group 7 and 8 contained a similar average number of target-site types though they have distinct average 3' UTR length. If the reduced abundance of miRNA targets simply results from the variation of 3' UTR length, one would expect that the group with longest 3' UTRs would show the lowest RR value. However, group 7, which contained the most target-site types, but not the longest 3' UTRs, showed the lowest RR value for most datasets (all except dataset 3, which contained a relatively small number of mRNAs for calculation). Therefore, the positive correlation cannot simply be explained by variations in 3' UTR length, but is more likely due to the increasing number of target-site types.
It is well known that cancer cells epigenetically silence a number of genes by CpG island methylation. We next asked if the global down-regulation of miRNA targets in prostate cancer was a side effect of gene silencing by CpG island methylation. We found that the miRNA targets had significantly more CpG islands (average: 1.55) than all mRNAs (average: 1.38, P < 0.01, Wilcoxon-Man-Whitney test). However, randomly chosen groups of non-miRNA-target mRNAs with the same (or more) number of CpG islands as the targets, did not exhibit the same global down-regulation as the miRNA targets (RR > 1, P > 0.05). Moreover, we did not find correlations between the 3' UTR length (or miRNA target-site types) and the number of CpG islands (P > 0.05, spearman correlation test). Thus, the global down-regulation of miRNA targets cannot simply be explained by gene silencing due to CpG islands methylation.
Reduced transcript levels of miRNA targets in androgen-independent prostate cancer when compared with androgen-dependent prostate cancer
Down-regulation of miRNA targets in other cancer types and predicted function of these protein-coding genes in prostate cancer
Biological processes of down-regulated miRNA targets and antitargets in prostate cancer
Macromolecule metabolic process
Regulation of biological process
Multicellular organismal development
KEGG pathway analysis of down-regulated miRNA targets and antitargets in prostate cancer
Regulation of actin cytoskeleton
Wnt signaling pathway
MAPK signaling pathway
Escherichia coli infection
To determine whether the global down-regulation of miRNA targets is common in human cancers, we extended the global analysis to four other prevalent cancers, including three solid tumours and one leukaemia (Table 1). Significant differences in transcript levels between miRNA targets and total mRNAs were observed for all four cancers [see Additional file 1]. Strikingly, the expression levels of miRNA targets were significantly lower in breast cancer, lung adenocarcinoma and acute myeloid leukaemia than in corresponding normal tissues (RR < 1, P < 0.05). In contrast, the comparison of hepatocellular carcinoma and non-tumour liver tissue yielded an opposite relationship between the abundance of miRNA target and total mRNAs (RR > 1, P < 0.05). In this analysis, an RR value greater than 1 indicated an increased abundance of miRNA targets when compared with all mRNAs. The globally increased abundance of miRNA targets may reflect different mechanisms of regulation in liver cancer. Our extended analysis indicates that the differences in transcript levels between miRNA targets and total mRNAs may be a common feature in human cancers.
Different expression levels between miRNA targets and total mRNAs have been uncovered in the comparison between mature tissues and embryos  and miRNAs have been suggested to confer precision and robustness to developmental processes. In this study, we initially reported that miRNA targets expressed less on a global scale than total mRNAs in prostate tumours, relative to benign prostate tissues. Analysis of the protein levels of miRNA targets suggests that the level of protein expression of miRNA targets may also be reduced, in agreement with a recent study which reported that the regulation of protein-coding genes by miRNAs was quite similar on the transcript and protein levels . Moreover, our data showed that the transcript abundance of the targets of androgen-repressed miRNAs was significantly lower than the abundance of the targets of all other miRNAs in androgen-independent prostate cancer. The abnormal expression pattern of miRNA targets was also seen in three other cancer types, suggesting that it may be a common feature of the human cancer transcriptome.
We also found a trend for an increased down-regulation of mRNAs with longer 3' UTRs and more target-site types, consistent with a recent study showing that proliferating cells express mRNAs with shortened 3' UTRs and fewer miRNA target sites . It has been reported that for proteins with more interacting partners, their genes tend to be regulated by more miRNA types [24, 25]. Genes with more interactions may require more elaborate regulation at the posttranscriptional level because unwanted output of these proteins may lead to a more severe fitness effect. Moreover, miRNAs have been proposed to primarily target downstream network components such as transcription factors . Disrupted expression of the highly regulated miRNA target genes may reflect the fact that the regulatory network in cancer cells departs from the normal regulatory routine presented in benign cells. The molecular mechanisms determining the intriguing expression patterns of miRNA targets in cancer cells presented in this study remain to be elucidated. Based on our analysis, there are three potential reasons discussed below.
First, the abnormal transcript abundance of miRNA targets may indicate a significant influence of miRNAs on the expression of their target genes in prostate tumours. This view is supported by three observations: (1) the targets of 36 up-regulated miRNAs made a strong contribution to the low expression of all miRNA targets; (2) there was a weak (but significant) negative correlation between the score of the differential expression of individual miRNAs (published by Volinia et al.) and the RR value of the miRNA's targets; and (3) the propensity for low expression increased with the number of target-site types embedded in the 3' UTRs of the miRNA targets, suggesting the possibility of synergistic regulation by multiple different miRNAs in prostate cancer. In general, lower transcript level is attributed to transcription inhibition or mRNA decay. Since miRNA target prediction relies strongly on the characteristics of 3' UTRs, translational control by 3' UTRs may play a role in the down-regulation of miRNA targets in prostate cancer. It has been demonstrated that miRNAs can promote rapid mRNA degradation by accelerating deadenylation  and that miRNAs are involved in AU-rich Element (ARE)-mediated mRNA instability . Therefore, the low expression of miRNA targets may result from the action of miRNA-mediated mRNA decay in prostate cancer. Up-regulation of miRNAs in prostate tumours is common [3, 4, 19] and is consistent with the known oncogenic activity of many miRNAs [28, 29]. It has been reported that Dicer and other genes involved in miRNA processing are up-regulated in prostate cancer , indicating that the prostate tumour is more efficient than normal prostate tissue at processing miRNA precursors into mature miRNAs. These observations support the idea that miRNAs may be up-regulated on a global scale in prostate cancer, consistent with the global down-regulation of their targets.
It should be noted that the global down-regulation of miRNA targets is an overall effect that does not negate the fact that some miRNA targets are up-regulated in prostate tumours. A recent miRNA profiling study showed a tumour gene signature that contains up-regulated and down-regulated miRNAs in prostate cancer . This study also showed that binding of miRNAs to 3' UTR sequences can lead to both degradation and accumulation of the targeted mRNA in cancer cells. In the correlation analysis between the expression level of individual miRNAs and the expression level of their putative targets, we confirmed this observation on a global scale. More specifically, both an inverse and a positive correlation could occur between a miRNA and its target mRNAs in prostate cancer cells. Since a miRNA can regulate multiple targets and a single mRNA can be targeted by several different miRNAs, the global down-regulation of miRNA targets may largely depend on the overall effect of miRNA regulation. The second potential reason is that the global down-regulation of miRNA targets is an overall effect that may depend on: (1) the reduction of mRNA expression that may be caused by the up-regulated miRNAs (such as miR-29a); (2) the decrease of mRNA sequestration that may be caused by the down-regulated miRNAs (such as let-7i); and (3) the moderate up-regulation of some targets of the down-regulated miRNAs (such as miR-125b). Furthermore, this potential reason can explain the fact that global up-regulation of miRNA targets was observed for hepatocellular carcinoma, a tissue which has roughly equal numbers of up and down-regulated miRNAs .
The third reason is related to a perplexing problem: several other miRNA profiling studies showed widespread down-regulation of miRNAs in prostate cancer [5–8], and some of the miRNAs reported to be up-regulated by Volinia et al. overlapped with some of the miRNAs reported to be down-regulated by another profiling study . If most miRNAs are truly down-regulated in prostate cancer, the global down-regulation of miRNA targets may not be causatively linked to the expression levels of the miRNAs themselves. As miRNA targets have relatively long 3' UTRs (with known functions in the stability, localisation, and translation of mRNA) and more CpG islands (which may be methylated in cancers) than non-targets, a third explanation is that the target genes that are down-regulated in prostate cancer are key cellular regulators and such key regulators are themselves highly regulated at multiple levels (transcriptionally and post-transcriptionally), including regulation by miRNAs.
There are several explanations for the discrepancies in miRNA profiling studies. First, Volinia et al.  and Ambs et al.  used total RNA while other studies [5–8] used purified small RNA samples (from 18 to 300 nt). Purification might introduce errors into miRNA expression comparisons. For example, there is presently no way to judge the different proportions of miRNAs within the pool of total RNAs . If cancer and benign tissue have different proportions of miRNA content, the validity of the analysis based on the fundamental assumption that same amount of miRNAs is extracted from the same amount of total RNAs is thrown into doubt. Second, it has been suggested that purification of the small RNA fraction could reduce nonspecific hybridisation to longer miRNA precursors. If there is a block in precursor miRNA processing in prostate cancer without a corresponding decrease in transcription, this could result in the inconsistency. A third explanation would be the differences in samples number. Volinia et al. and Ambs et al. analysed > 50 prostate cancer samples while other studies did not reach this size. Cancer is a heterogeneous disease, and the heterogeneity of tumour samples might contribute substantially to the results. It is not surprising that miRNA expression profiles published by different researchers are inconsistent, because miRNA profiling technology is still in its infancy. For example, researchers generally adopt "tried-and-true" methodologies from cDNA microarray technology for miRNA expression analysis, but the relatively small number of probes on miRNA microarrays may render these high-density approaches ineffective.
As to the basis of our investigation, the gene expression microarray and miRNA target prediction data have proved to be useful for gaining biological knowledge [13, 22, 24–26]. Although these datasets are far from being complete and may contain noise, it is unlikely that these flaws could totally distort the results. Since consistent results were seen across various datasets generated by independent groups, the noise of microarray data and false positives in miRNA target predictions appear to have no serious effects on our study. Furthermore, the overall significances inferred from thousands of mRNAs would be strong enough to reflect real biology. The strength of our global analysis lies in the noise reduction effect, as well as the identification of general trends of miRNA target expression that would not have been discovered by individual investigation of single miRNA targets. Cancer is an extremely complex and heterogeneous disease . It should be noted that our data did not conclusively distinguish among the three possible mechanisms discussed above, and the detailed molecular mechanisms responsible for the abnormal expression of miRNA targets remain to be thoroughly elucidated. Future experiments or large microarray studies are needed to clarify the possible mechanisms.
In conclusion, our global gene expression analysis, along with further investigations, suggests that miRNA targets have significantly reduced transcript abundance in prostate cancer, when compared with the combined pool of total mRNAs. The abnormal expression patterns of miRNA targets could be a common feature of the human cancer transcriptome. These observations raise the possibility that miRNA may have global functions in human prostate cancer. Our study may help to shed new light on miRNA functions in cancer transcriptomics, when unprecedented opportunities to study the regulatory control mediated by miRNAs are given by the accumulation of cDNA, miRNA expression and proteomic datasets.
The gene expression microarray datasets used in this study are listed in Table 1, including five datasets in human prostate cancer and four in other cancer types (breast cancer, lung adenocarcinoma, acute myeloid leukaemia and liver cancer) [33–36]. We also used one immunobloting proteomic dataset in prostate cancer. Gene expression datasets were of two general types, two channel ratio data (cDNA datasets) and single channel intensity data (Affymetrix datasets), and were generally given in a single matrix file format. All gene expression datasets were normalized by the authors of these studies. Probe IDs were converted to RefSeq mRNA IDs using ID converter , if necessary. We used two complete lists of human miRNA targets published by Lewis et al. and Krek et al. These miRNA target prediction datasets were downloaded from the most recently updated websites.
Determining the mRNA groups with high and low expression in cancers relative to normal tissues
We reviewed the samples profiled for each of the gene expression datasets and chose the samples of classes of interest for further analysis. For each of the two classes e.g. cancer versus normal, the probe sets with absent calls (Affymetrix) or missing values (cDNA) in excess of 50% of the samples were filtered out. For each probe, we first averaged across sample replicates, then directly compared the median expression values between the two classes (cancer and normal). After that, we averaged the ratios (fold changes) for probes of the same RefSeq transcripts in each dataset and determined whether a mRNA had a high or low expression level (Method a). In dataset 1 and 2, where the expression measurements of matched samples (prostate cancers and benign prostate samples from same patients) were provided, another method (Method b) was used to rule out the bias of single method. We determined if a mRNA altered for each patient by computing the expression ratios of each patient, and only those genes showing alterations in most (based on different cut-offs of median ratio of paired samples) patients were considered to be altered. In dataset 2, only the "well-measured genes" defined by the author were included in our analysis. Moreover, in order to rule out the bias of single cut-off (or threshold) for identifying mRNAs with low and high expression, we also chose three different cut-off values for each comparison (High, Medium and Low [see Additional file 1]). For datasets 5–7 of three other cancer types, we directly downloaded the over- and under-expression gene lists from Oncomine database .
Calculation of R and RR values and resampling statistical tests
Here we modified a method from Yu et al. 2007 . After obtaining the mRNAs with high and low expression in cancers relative to normal tissues, we counted the number of miRNA targets that fell in the high expression group (N> Benign) and divided it by the amount of targets in the low expression group (N< Benign). The obtained ratio was named as Rmir (Rmir = N> Benign/N< Benign). As a control, the same calculations for all mRNAs were undertaken and the ratio Rtotal was obtained. The ratio of Rmir to Rtotal was computed and termed as RR (RR = Rmir/Rtotal). The RR value is a surrogate for an increased or decreased abundance of miRNA targets relative to total mRNAs. An RR value less than 1 indicates that the targets prefer a lower expression. We undertook a resampling test to judge the statistical significance of the global observation. In each test, we randomly picked the same number of mRNAs as the number of miRNA targets from total altered mRNAs. We then calculated the ratio of high expression mRNAs to low expression mRNAs in this random sub-pool and termed it as Rrandom. The randomization tests were performed 5000 times and the number of times (n) was counted when Rrandom > Rmir. P-value was defined as n/5000. If P-value < 0.05, we determined the expression level of miRNA targets was significantly lower than that of total mRNAs. We used two miRNA target predictions for this analysis and performed correlation analysis of the RR values obtained from each of the two target predictions.
Correlation analysis between the expression levels of individual miRNAs and those of their putative targets
For the dataset containing expression measurements of both miRNAs and mRNAs in ten prostate tumours and ten corresponding surrounding non-tumour tissues , we calculated the log expression ratios of paired samples for each miRNA and each mRNA. Then the Pearson correlation coefficients between each miRNA and each mRNA were calculated using the expression ratios. We also determined the global density distribution of the Pearson correlation coefficients between the miRNA of interest and either all mRNAs or the predicted targets of the miRNA.
Analysis of the relationship between the RR values and 3' UTR length and the number of predicted miRNA target-site types
The number of target-site types of each PicTar target mRNA was calculated. Since the underlying distributions of the numbers of miRNA target site types and RR values were not normal, Spearman's rank correlation test was used to determine the relations between these two variables. We extracted information of 3' UTR length of each mRNA from the UTResource database . The human CpG islands data were extracted from Human CpG Island database .
GO term analysis and KEGG pathway analysis
To study the function of down-regulated miRNA targets, we used the GO Term Mapper Web server . We used default GOA slim file (a list of general GO terms) for annotating the down-regulated mRNAs. The KEGG database  batch entry allowed us to evaluate the large set of down-regulated miRNA targets and antitargets. In order to evaluate the enrichment level of mRNAs to a specific biological process or pathway, we used the Hyper-Geometric (HG) distribution to calculate the enrichment P value. A more detailed explanation of this distribution was previously described .
We are grateful to two anonymous reviewers for their thoughtful criticisms, comments and suggestions on early version of the manuscript. We also thank Ms. Zhaorong Ma for copyediting of the manuscript. This work is supported by a grant 2006AA02Z324 from the National High Technology Research and Development Program of China (863 Program). This work is also supported by Shanghai Leading Academic Discipline Project, Project Number: B111.
- Wu L, Fan J, Belasco JG: MicroRNAs direct rapid deadenylation of mRNA. Proc Natl Acad Sci USA. 2006, 103 (11): 4034-4039. 10.1073/pnas.0510928103.PubMed CentralView ArticlePubMedGoogle Scholar
- Lewis BP, Burge CB, Bartel DP: Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell. 2005, 120 (1): 15-20. 10.1016/j.cell.2004.12.035.View ArticlePubMedGoogle Scholar
- Esquela-Kerscher A, Slack FJ: Oncomirs – microRNAs with a role in cancer. Nat Rev Cancer. 2006, 6 (4): 259-269. 10.1038/nrc1840.View ArticlePubMedGoogle Scholar
- Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, Visone R, Iorio M, Roldo C, Ferracin M, et al: A microRNA expression signature of human solid tumours defines cancer gene targets. Proc Natl Acad Sci USA. 2006, 103 (7): 2257-2261. 10.1073/pnas.0510565103.PubMed CentralView ArticlePubMedGoogle Scholar
- Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, Peck D, Sweet-Cordero A, Ebert BL, Mak RH, Ferrando AA, et al: MicroRNA expression profiles classify human cancers. Nature. 2005, 435 (7043): 834-838. 10.1038/nature03702.View ArticlePubMedGoogle Scholar
- Mattie MD, Benz CC, Bowers J, Sensinger K, Wong L, Scott GK, Fedele V, Ginzinger D, Getts R, Haqq C: Optimized high-throughput microRNA expression profiling provides novel biomarker assessment of clinical prostate and breast cancer biopsies. Mol Cancer. 2006, 5: 24-10.1186/1476-4598-5-24.PubMed CentralView ArticlePubMedGoogle Scholar
- Porkka KP, Pfeiffer MJ, Waltering KK, Vessella RL, Tammela TL, Visakorpi T: MicroRNA expression profiling in prostate cancer. Cancer Res. 2007, 67 (13): 6130-6135. 10.1158/0008-5472.CAN-07-0533.View ArticlePubMedGoogle Scholar
- Ozen M, Creighton CJ, Ozdemir M, Ittmann M: Widespread deregulation of microRNA expression in human prostate cancer. Oncogene. 2008, 27 (12): 1788-1793. 10.1038/sj.onc.1210809.View ArticlePubMedGoogle Scholar
- Krek A, Grun D, Poy MN, Wolf R, Rosenberg L, Epstein EJ, MacMenamin P, da Piedade I, Gunsalus KC, Stoffel M, et al: Combinatorial microRNA target predictions. Nat Genet. 2005, 37 (5): 495-500. 10.1038/ng1536.View ArticlePubMedGoogle Scholar
- Johnson SM, Grosshans H, Shingara J, Byrom M, Jarvis R, Cheng A, Labourier E, Reinert KL, Brown D, Slack FJ: RAS is regulated by the let-7 microRNA family. Cell. 2005, 120 (5): 635-647. 10.1016/j.cell.2005.01.014.View ArticlePubMedGoogle Scholar
- Farh KK, Grimson A, Jan C, Lewis BP, Johnston WK, Lim LP, Burge CB, Bartel DP: The widespread impact of mammalian MicroRNAs on mRNA repression and evolution. Science. 2005, 310 (5755): 1817-1821. 10.1126/science.1121158.View ArticlePubMedGoogle Scholar
- Stark A, Brennecke J, Bushati N, Russell RB, Cohen SM: Animal MicroRNAs confer robustness to gene expression and have a significant impact on 3'UTR evolution. Cell. 2005, 123 (6): 1133-1146. 10.1016/j.cell.2005.11.023.View ArticlePubMedGoogle Scholar
- Sood P, Krek A, Zavolan M, Macino G, Rajewsky N: Cell-type-specific signatures of microRNAs on target mRNA expression. Proc Natl Acad Sci USA. 2006, 103 (8): 2746-2751. 10.1073/pnas.0511045103.PubMed CentralView ArticlePubMedGoogle Scholar
- Selbach M, Schwanhausser B, Thierfelder N, Fang Z, Khanin R, Rajewsky N: Widespread changes in protein synthesis induced by microRNAs. Nature. 2008, 455 (7209): 58-63. 10.1038/nature07228.View ArticlePubMedGoogle Scholar
- Yu YP, Landsittel D, Jing L, Nelson J, Ren B, Liu L, McDonald C, Thomas R, Dhir R, Finkelstein S, et al: Gene expression alterations in prostate cancer predicting tumour aggression and preceding development of malignancy. J Clin Oncol. 2004, 22 (14): 2790-2799. 10.1200/JCO.2004.05.158.View ArticlePubMedGoogle Scholar
- Lapointe J, Li C, Higgins JP, Rijn van de M, Bair E, Montgomery K, Ferrari M, Egevad L, Rayford W, Bergerheim U, et al: Gene expression profiling identifies clinically relevant subtypes of prostate cancer. Proc Natl Acad Sci USA. 2004, 101 (3): 811-816. 10.1073/pnas.0304146101.PubMed CentralView ArticlePubMedGoogle Scholar
- Dhanasekaran SM, Barrette TR, Ghosh D, Shah R, Varambally S, Kurachi K, Pienta KJ, Rubin MA, Chinnaiyan AM: Delineation of prognostic biomarkers in prostate cancer. Nature. 2001, 412 (6849): 822-826. 10.1038/35090585.View ArticlePubMedGoogle Scholar
- Varambally S, Yu J, Laxman B, Rhodes DR, Mehra R, Tomlins SA, Shah RB, Chandran U, Monzon FA, Becich MJ, et al: Integrative genomic and proteomic analysis of prostate cancer reveals signatures of metastatic progression. Cancer Cell. 2005, 8 (5): 393-406. 10.1016/j.ccr.2005.10.001.View ArticlePubMedGoogle Scholar
- Ambs S, Prueitt RL, Yi M, Hudson RS, Howe TM, Petrocca F, Wallace TA, Liu CG, Volinia S, Calin GA, et al: Genomic profiling of microRNA and messenger RNA reveals deregulated microRNA expression in prostate cancer. Cancer Res. 2008, 68 (15): 6162-6170. 10.1158/0008-5472.CAN-08-0144.PubMed CentralView ArticlePubMedGoogle Scholar
- Moore MJ: From birth to death: the complex lives of eukaryotic mRNAs. Science. 2005, 309 (5740): 1514-1518. 10.1126/science.1111443.View ArticlePubMedGoogle Scholar
- Jiang M, Li M, Fu X, Huang Y, Qian H, Sun R, Mao Y, Xie Y, Li Y: Simultaneously detection of genomic and expression alterations in prostate cancer using cDNA microarray. Prostate. 2008, 68 (14): 1496-1509. 10.1002/pros.20756.View ArticlePubMedGoogle Scholar
- Yu Z, Jian Z, Shen SH, Purisima E, Wang E: Global analysis of microRNA target gene expression reveals that miRNA targets are lower expressed in mature mouse and Drosophila tissues than in the embryos. Nucleic Acids Res. 2007, 35 (1): 152-164. 10.1093/nar/gkl1032.PubMed CentralView ArticlePubMedGoogle Scholar
- Sandberg R, Neilson JR, Sarma A, Sharp PA, Burge CB: Proliferating cells express mRNAs with shortened 3' untranslated regions and fewer microRNA target sites. Science. 2008, 320 (5883): 1643-1647. 10.1126/science.1155390.PubMed CentralView ArticlePubMedGoogle Scholar
- Liang H, Li WH: MicroRNA regulation of human protein protein interaction network. Rna. 2007, 13 (9): 1402-1408. 10.1261/rna.634607.PubMed CentralView ArticlePubMedGoogle Scholar
- Cui Q, Yu Z, Pan Y, Purisima EO, Wang E: MicroRNAs preferentially target the genes with high transcriptional regulation complexity. Biochem Biophys Res Commun. 2007, 352 (3): 733-738. 10.1016/j.bbrc.2006.11.080.View ArticlePubMedGoogle Scholar
- Cui Q, Yu Z, Purisima EO, Wang E: Principles of microRNA regulation of a human cellular signaling network. Mol Syst Biol. 2006, 2: 46-10.1038/msb4100089.PubMed CentralView ArticlePubMedGoogle Scholar
- Jing Q, Huang S, Guth S, Zarubin T, Motoyama A, Chen J, Di Padova F, Lin SC, Gram H, Han J: Involvement of microRNA in AU-rich element-mediated mRNA instability. Cell. 2005, 120 (5): 623-634. 10.1016/j.cell.2004.12.038.View ArticlePubMedGoogle Scholar
- Landais S, Landry S, Legault P, Rassart E: Oncogenic potential of the miR-106-363 cluster and its implication in human T-cell leukemia. Cancer Res. 2007, 67 (12): 5699-5707. 10.1158/0008-5472.CAN-06-4478.View ArticlePubMedGoogle Scholar
- Ma L, Teruya-Feldstein J, Weinberg RA: Tumour invasion and metastasis initiated by microRNA-10b in breast cancer. Nature. 2007, 449 (7163): 682-688. 10.1038/nature06174.View ArticlePubMedGoogle Scholar
- Li W, Xie L, He X, Li J, Tu K, Wei L, Wu J, Guo Y, Ma X, Zhang P, et al: Diagnostic and prognostic implications of microRNAs in human hepatocellular carcinoma. Int J Cancer. 2008, 123 (7): 1616-1622. 10.1002/ijc.23693.View ArticlePubMedGoogle Scholar
- Davison TS, Johnson CD, Andruss BF: Analyzing micro-RNA expression using microarrays. Methods Enzymol. 2006, 411: 14-34. 10.1016/S0076-6879(06)11002-2.View ArticlePubMedGoogle Scholar
- Wang E, Lenferink A, O'Connor-McCourt M: Cancer systems biology: exploring cancer-associated genes on cellular networks. Cell Mol Life Sci. 2007, 64 (14): 1752-1762. 10.1007/s00018-007-7054-6.View ArticlePubMedGoogle Scholar
- Richardson AL, Wang ZC, De Nicolo A, Lu X, Brown M, Miron A, Liao X, Iglehart JD, Livingston DM, Ganesan S: X chromosomal abnormalities in basal-like human breast cancer. Cancer Cell. 2006, 9 (2): 121-132. 10.1016/j.ccr.2006.01.013.View ArticlePubMedGoogle Scholar
- Beer DG, Kardia SL, Huang CC, Giordano TJ, Levin AM, Misek DE, Lin L, Chen G, Gharib TG, Thomas DG, et al: Gene-expression profiles predict survival of patients with lung adenocarcinoma. Nat Med. 2002, 8 (8): 816-824.PubMedGoogle Scholar
- Andersson A, Ritz C, Lindgren D, Eden P, Lassen C, Heldrup J, Olofsson T, Rade J, Fontes M, Porwit-Macdonald A, et al: Microarray-based classification of a consecutive series of 121 childhood acute leukemias: prediction of leukemic and genetic subtype as well as of minimal residual disease status. Leukemia. 2007, 21 (6): 1198-1203. 10.1038/sj.leu.2404688.View ArticlePubMedGoogle Scholar
- Chen X, Cheung ST, So S, Fan ST, Barry C, Higgins J, Lai KM, Ji J, Dudoit S, Ng IO, et al: Gene expression patterns in human liver cancers. Mol Biol Cell. 2002, 13 (6): 1929-1939. 10.1091/mbc.02-02-0023..PubMed CentralView ArticlePubMedGoogle Scholar
- Clone/Gene ID Converter. [http://idconverter.bioinfo.cnio.es/]
- Rhodes DR, Kalyana-Sundaram S, Mahavisno V, Varambally R, Yu J, Briggs BB, Barrette TR, Anstet MJ, Kincead-Beal C, Kulkarni P, et al: Oncomine 3.0: genes, pathways, and networks in a collection of 18,000 cancer gene expression profiles. Neoplasia. 2007, 9 (2): 166-180. 10.1593/neo.07112.PubMed CentralView ArticlePubMedGoogle Scholar
- UTResource Database. [http://www.ba.itb.cnr.it/UTR/UTRHome.html]
- Human CpG Island Database release 3.0. [ftp://ftp.ebi.ac.uk/pub/databases/cpgisle]
- GO TERM MAPPER. [http://go.princeton.edu/cgi-bin/GOTermMapper]
- KEGG Database. [http://www.genome.jp/kegg]
- Rokhlenko O, Wexler Y, Yakhini Z: Similarities and differences of gene expression in yeast stress conditions. Bioinformatics. 2007, 23 (2): e184-190. 10.1093/bioinformatics/btl308.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.