Int. J. Mol. Sci. 2013, 14(6), 11560-11606; doi:10.3390/ijms140611560

MicroRNA-Regulated Protein-Protein Interaction Networks and Their Functions in Breast Cancer
Chia-Hsien Lee 1,, Wen-Hong Kuo 2,, Chen-Ching Lin 1, Yen-Jen Oyang 1, Hsuan-Cheng Huang 3,* and Hsueh-Fen Juan 1,4,*
Graduate Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei 106, Taiwan; E-Mails: (C.-H.L.); (C.-C.L.); (Y.-J.O.)
Department of Physiology, College of Medicine, National Taiwan University, Taipei 100, Taiwan; E-Mail:
Institute of Biomedical Informatics and Center for Systems and Synthetic Biology, National Yang-Ming University, Taipei 112, Taiwan
Institute of Molecular and Cellular Biology and Department of Life Science, National Taiwan University, Taipei 106, Taiwan
These authors contributed equally to this work.
Authors to whom correspondence should be addressed; E-Mails: (H.-C.H.); (H.-F.J.); Tel.: +886-2-2826-7357 (H.-C.H.); +886-2-3366-4536 (H.-F.J.); Fax: +886-2-2820-2508 (H.-C.H.); +886-2-2367-3374 (H.-F.J.).
Received: 17 April 2013; in revised form: 21 May 2013 / Accepted: 22 May 2013 /
Published: 30 May 2013


: MicroRNAs, which are small endogenous RNA regulators, have been associated with various types of cancer. Breast cancer is a major health threat for women worldwide. Many miRNAs were reported to be associated with the progression and carcinogenesis of breast cancer. In this study, we aimed to discover novel breast cancer-related miRNAs and to elucidate their functions. First, we identified confident miRNA-target pairs by combining data from miRNA target prediction databases and expression profiles of miRNA and mRNA. Then, miRNA-regulated protein interaction networks (PINs) were constructed with confident pairs and known interaction data in the human protein reference database (HPRD). Finally, the functions of miRNA-regulated PINs were elucidated by functional enrichment analysis. From the results, we identified some previously reported breast cancer-related miRNAs and functions of the PINs, e.g., miR-125b, miR-125a, miR-21, and miR-497. Some novel miRNAs without known association to breast cancer were also found, and the putative functions of their PINs were also elucidated. These include miR-139 and miR-383. Furthermore, we validated our results by receiver operating characteristic (ROC) curve analysis using our miRNA expression profile data, gene expression-based outcome for breast cancer online (GOBO) survival analysis, and a literature search. Our results may provide new insights for research in breast cancer-associated miRNAs.
miRNA; breast cancer; protein interaction network; functional analysis

1. Introduction

Breast cancer is a global health threat for women. According to a 2008 survey [1], breast cancer was the leading cause of cancer deaths in women. Our knowledge of possible risk factors has led to developments in diagnostic methods, drugs, and surgery procedures for treatment [2,3]; however, the details of breast carcinoma progression, and perhaps most importantly, how to cure breast cancer, remain elusive.

Previous research has identified a number of risk factors for breast cancer. Early menarche, late menopause, obesity, late first full pregnancy, and hormone replacement therapy were considered as high risk factors for breast cancer [2]. Breast cancer risk has also been reported to be related to fat intake in diets rich in red meats and high-fat dairy foods [3].

MicroRNAs (miRNAs) [4], are short endogenous non-coding RNAs which are able to regulate gene expression. After miRNA precursors are transcribed from the genome or generated from spliceosomes, they are exported to the cytoplasm and further processed by the Dicer complex [5]. The mature miRNA is then bound to Argonaute protein, forming a miRNA-protein complex known as the RNA-induced silencing complex (RISC) [6], miRNP, or RNAi (RNA interference) enzyme complex [7,8]. The RISC has been reported to down-regulate target genes by translational repression [9] or mRNA cleavage [10].

Like other protein-based regulators, miRNAs have been associated with cancer. Calin et al., reported that miR-15 and miR-16 were deleted in leukemia [11], which was believed to be one of the earliest reports associating miRNAs with cancer [12]. After this report, many miRNAs were found to act as tumor suppressors or oncogenes (also known as oncomirs). For example, miR-21 was identified as an oncomir in hepatocellular cancer [13], breast cancer [14], and kidney cancer [15]. On the other hand, let-7c was found to be a tumor suppressor in prostate cancer [16], and miR-181a was reported as a tumor suppressor in glioma [17]. Further, miR-125b [18], and miR-145 [19] were identified as tumor suppressors in breast cancer, and miR-125a was found to repress tumor growth in breast cancer [20]. Thus, it is highly likely that miRNAs play an important role in breast cancer.

Since miRNA functions by regulating its target genes, we may deduce the effects of miRNAs by analyzing their regulated networks. To use such a method to elucidate miRNA functions, targets of miRNAs should be deduced. Currently, predictions in most target prediction database are based on sequence and statistical methods [21]. For example, in TargetScan, seed base pairing, target site context, conservation of target site and miRNA, and site accessibility are considered in the prediction process [22].

Another method to elucidate miRNA targets is to integrate expression profiles of miRNA and mRNA. In the work of Huang et al.[23], a Bayesian-based algorithm, GenMiR++, was developed to predict possible targets of 104 miRNAs in humans. They also verified their results by RT-PCR and microarray experiments. However, the power of other sequence-based target prediction algorithms was not utilized in their work.

It is also possible to combine sequence-based target prediction and expression-based target prediction methods. By integrating expression data into sequence-based predictions, possible false positives can be reduced. Previously, miRNA-mRNA interactions were explored with splitting-averaging Bayesian networks [24]. In that work, expression profiles of miRNA and mRNA from public databases, miRNA target prediction databases, and miRNA sequence information were integrated together to discover miRNA-mRNA interaction networks.

Here, we combined expression profiles of miRNA and mRNA, and three target prediction databases, TargetScan, PicTar and miRanda, to obtain confident miRNA-mRNA relationships and construct miRNA-regulated protein-protein networks for breast cancer. Furthermore, we explored the functions of the miRNAs by inspecting the underlying protein interaction networks (PINs) of the miRNAs with functional enrichment analysis. This method, as described in Figure 1, was used to elucidate the functions of gastric cancer-related miRNAs in our previous work [25]. In that study, a gastric cancer-associated miRNA, miR-148a, was identified and validated as being involved in tumor proliferation, invasion, migration, and the survival rate of the patients. By using a similar method, we aimed to elucidate breast cancer-related miRNA-regulated PINs and their functions.

2. Results and Discussion

To construct miRNA-regulated PINs, differentially expressed miRNAs and genes from the dataset from Farazi et al.[26] were extracted following proper processing of the expression profiles. From our selected public miRNA dataset, we found 89 down-regulated miRNAs (93 prior to fold-change filtering) and only 1 up-regulated miRNA (Table S1). In gene expression dataset GSE29174, we found a total of 1268 down-regulated genes and 587 up-regulated genes before applying the fold change filter. There were 726 down-regulated genes (Table S2) and 437 up-regulated genes (Table S3) after significantly and differentially expressed genes were filtered by fold change (fold change >2).

From the results of SAM analysis, we identified some well-known breast cancer-related miRNAs (Table S1). For example, miR-214-3p [27] and miR-335-5p [28] have been previously reported to be down-regulated in breast cancer. Let-7c was found to be down-regulated in this work, while let-7a, another member of the let-7 family, was found to be down-regulated in another work [29]. MicroRNAs of the let-7 family were also reportedly down-regulated in several types of cancer [30]. We also found that miR-21-5p, the sole up-regulated miRNA in our list, was also previously found to be up-regulated [14,31]. However, changes in the expression of most of the miRNAs in our down-regulated list have not been reported in the literature. Therefore, we could not rule out the possibility that these miRNAs were novel breast cancer-related miRNAs. There are also some well-known miRNAs not presented in our list (for such a list, one may see [3234]). The reason that some known miRNAs, for example, miR-19a, miR-155 and miR-205, did not show up in our result might be that we used a very stringent threshold (described in Experimental Section) when selecting differentially expressed miRNAs for PIN construction.

Since the miRNAs of the miRNA-regulated PINs were differentially expressed between normal and tumor tissues, and we identified some cancer-related functions in our functional enrichment analysis, the miRNAs may potentially be useful diagnostic markers for breast cancer. To verify this, we applied ROC curve analysis on the miRNA expression profile that was not used in constructing the miRNA-regulated PINs. Notably, our results (Figures 2 and S1 and Table S4) showed that let-7c (Figure 2), miR-497-5p, miR-125b-5p, and some other miRNAs of miRNA-regulated PINs, performed well when used as breast cancer diagnostic markers.

Following elucidation of differentially expressed miRNAs and genes, miRNA-regulated PINs could then be constructed. We identified and constructed partial networks, containing the miRNA and its direct target, with the differentially expressed miRNAs and genes as described in the “Experimental Section”. We then extended the network by appending known interactions from the HPRD database. Finally, 18 miRNA-regulated PINs were constructed by the steps described above (Figure 3, Figures S2–S13, and Table S5).

After construction of the 18 PINs was completed, we observed that the sizes of the PINs were not similar: some of the miRNAs seemed to regulate larger sized PINs, while other miRNAs affected only a small number of genes. Small miRNA-regulated PINs may be caused by the strict q-value threshold set during SAM analysis, the processing steps performed on the target prediction databases discussed previously, and possibly by lack of protein-protein interaction data for some proteins in the HPRD. Although the HPRD may be considered the most comprehensive source of protein-protein interaction data [35], some proteins may not have been considered and researched by other investigators, and therefore, interaction data for those proteins would not be included in the HPRD. However, it may be true that some of the miRNA-regulated PINs were small in breast cancer, since the construction of the PINs were based on differentially expressed miRNAs and genes between normal tissues and tumor samples, and those miRNAs with small PINs may not be as important as others with larger PINs.

To elucidate the functions of a miRNA with a regulated PIN, GO enrichment analysis was applied to the miRNA-regulated PINs. We did not consider the miRNAs with ≤5 genes in their regulated PINs, and some of the miRNA-regulated PINs had no enriched functions using the defined threshold, FDR < 0.0001. To exclude GO terms that describe a broad range of concepts, we only included high level GO terms, i.e., larger than 5.

The results of the GO enrichment analysis for let-7c related to cancer are listed in Table 1. (Results of all miRNA-regulated PINs were in Table S6). We defined a GO term as cancer-related if a GO term contained “cell proliferation”, “cell death”, “apoptosis”, “signaling”, “microtubule”, and “actin”. We noted that 7 miRNAs had enriched GO terms related to apoptosis, cell death, and cell proliferation, i.e., miR-520d-3p, miR-497-5p, miR-125b-5p, miR-21-5p, miR-31-5p, let-7c, and miR-125-5p. Further, some miRNA-regulated PINs may have functions other than cell survival. For example, the nerve growth factor receptor pathway was enriched in miR-regulated PINs of miR-520d-3p, miR-497-5p, miR-125a-5p, miR-125b-5p, and miR-31-5p, and the epidermal growth factor receptor pathway was enriched in miR-regulated PINs of miR-520d-3p, miR-21-5p, and miR-497-5p. Most of the miRNAs had been previously described and were known to be implicated in breast cancer. Let-7c was not only likely to be down-regulated in breast cancer [29], but was also found to be a tumor suppressor in prostate cancer [16]. Another reported tumor suppressor was miR-125b-5p, which was found to be down-regulated in breast tumor tissue [18], and this finding was consistent with our functional enrichment results. The miR-125a-5p-regulated PIN was found to be able to inhibit apoptosis and regulate epithelial cell proliferation, and has been reported to repress cell growth [36].

We also found that some breast cancer-specific functions were enriched in our results. For example, in the miR-497-5p-regulated PIN, the term “androgen receptor signaling pathway” was enriched. Although it is not clear whether androgens are related to breast cancer, androgen receptors are known to be up-regulated in breast cancer and related to node invasiveness [37].

To further verify the results of the enriched cancer-related functions, we used GOBO for survival analysis. Our hypothesis is that expression of genes annotated with enriched cancer-related terms may be related to survival outcome of patients. With exception to some cell death/proliferation-related terms, it is already known that some pathways or functions are also related to clinical outcomes. For example, cell proliferation-related GO terms have a high probability of affecting survival of cancer tissues, and patient outcome may worsen if cancer tissue survives. In addition, some signaling pathways were known to enhance invasiveness, migration abilities, or were associated with reduced patient survival. For example, the BMP signaling pathway is known to confer various tumor cells with enhanced migration and invasion abilities [38], and nerve growth factor receptor (NGFR) was found to be associated with overall survival of breast cancer [38]. Furthermore, Toll-like receptor 4 has been reported to promote adhesion and invasive migration in breast cancer [39]. Finally, the cytoskeleton plays an important role in regulating cell motility in all cells. Actin filaments are known to participate in the invasive migration of cancer cells [40]. Since some of these functions were present in our enriched terms, we wished to test if the expression of a gene set annotated to the cancer-related enriched terms in the PIN would be related to clinical outcome of patients.

As shown in Figure 4, Figure S14 and Table S7, only some of the enriched terms were significantly associated with clinical outcome. This may be because the changes in these key genes occurred at the protein level, such as in protein expression or even post-translational modification; therefore, mRNA expression-centric tools like GOBO cannot explore association of such genes to clinical outcomes. Alternatively, it is possible the miRNA did not regulate the whole pathway, or the miRNA did not target the key part of the pathway directly, and thus the clinical outcome of gene sets of the enriched GO terms in this condition cannot be determined. However, some functions associated with clinical outcomes were observed. For example, proteins annotated with the terms “microtubule cytoskeleton”, “negative regulation of programmed cell death”, and “negative regulation of cell death” in the let-7c-regulated PIN were related to 10 year survival rate of patients, as reported by GOBO (Figure 4). Also, the enriched term “regulation of epithelial cell proliferation” for both miR-125a-5p and miR-125b-5p were found to be associated with the 10-year survival rate of patients. Therefore, these results further supported the GO enrichment analysis discussed previously.

3. Experimental Section

3.1. miRNA Microarray Experiments

We performed a miRNA microarray to obtain the expression profiles for receiver operating characteristic (ROC) curve analysis. This dataset was deposited in Gene Expression Omnibus (GEO,, Series Accession: GSE45666). In total, there were 15 normal samples and 101 tumor samples in the expression profile. Detailed pathophysiogical characteistics of these samples were in Table S8 and GSE45666. All human tissue samples collected from breast cancer patients were approved and human subject confidentiality was protected by the Institute Review Board of National Taiwan University Hospital (IRB, 20071211R).

Total RNA was extracted from tissues collected from the patients using Trizol® Reagent (Invitrogen, Carlsbad, CA, USA, USA) according to the manufacturer’s protocol. Purified RNA was quantified at OD260nm by using a ND-1000 spectrophotometer (Nanodrop Technology, Wilmington, DE, USA) and qualitated by using a Bioanalyzer 2100 with the RNA 6000 Nano LabChip kit (Agilent Technologies, Santa Clara, CA, USA).

After RNA extraction, 100 ng of total RNA was dephosphorylated and labeled with pCp-Cy3 using the Agilent miRNA Complete Labeling and Hyb Kit in conjunction with the microRNA Spike-In kit (Agilent Technologies, Santa Clara, CA, USA). Briefly, 2X hybridization buffer (Agilent Technologies) was added to the labeled mixture to a final volume of 45 μL. The mixture was heated for 5 min at 100 °C and immediately cooled to 0 °C. Each 45 μL sample was hybridized onto an Agilent human miRNA Microarray Release 12.0, 8 × 15 K (Agilent Technologies) at 55 °C for 20 h. After hybridization, slides were washed for 5 min in Gene Expression Wash Buffer 1 at room temperature, then for 5 min in Gene Expression Wash Buffer 2 at 37 °C. Slides were scanned on an Agilent microarray scanner (Agilent Technologies, model G2505C) at 100% and 5% sensitivity settings. Feature Extraction (Agilent Technologies) software version (Agilent Technologies, Santa Clara, CA, USA) was used for image analysis.

3.2. mRNA Expression Profiles

For miRNA-regulated protein-protein interaction network construction, the mRNA expression profile was fetched from GEO (Series Accession: GSE29174). This dataset was produced by Farazi et al. [26], which was the only dataset publicly available with large size of tumor samples and reasonably sized normal samples. In total, 161 clinical samples were collected from breast cancer patients by biopsy: 110 invasive ductal carcinoma (IDC), 11 normal, 17 ductal carcinoma in situ (DCIS), 1 mucinous A, 8 atypical medullary, 4 apocrine, 8 metaplastic, and 2 adenoid, as classified by Farazi et al. The 110 IDC samples were classified as the tumor group and the 11 normal samples were classified as the normal group in this study.

3.3. miRNA Expression Profiles

The miRNA expression profile used in this work was fetched from Table S4A of the work of Farazi et al. [26]. There were 189 samples in this dataset, with 6 of them from cell lines, and another 183 samples from patient tissues. In the 183 clinical samples collected from breast cancer patients by biopsy in this dataset, there were 128 IDC, 11 normal, 18 DCIS, 1 mucinous A, 8 atypical medullary, 4 apocrine, 9 metaplastic, and 2 adenoid cases, as classified by Farazi et al. The 128 IDC samples were classified as the tumor group and the 11 normal samples as were classified as the normal group.

3.4. Data Analysis

The overall workflow design was similar to the previous work of Tseng et al. [25] (see Figure 1). However, we applied the workflow on breast cancer expression profiles instead of gastric cancer as in the work of Tseng et al. Additionally, we used 3 miRNA target prediction databases, TargetScan (v6.0) [22,41,42], PicTar [43,44], and miRanda (release August 2010) [45] here, while only TargetScan was used in the previous study.

To construct the networks, we first elucidated differentially expressed miRNAs and mRNAs from published datasets. Both of miRNA (generated with miRNA-Seq technique) and mRNA array expression data were produced by Farazi et al.[26] from the same batch of clinical tumor samples. To obtain a list of differentially expressed genes and miRNAs between normal and tumor groups, we used significance analysis of microarrays (SAM) [46] implemented in R package samr (version 2.0; Stanford University, Stanford, CA, USA). We set the false discovery rate (FDR) as ≤0.0001%, fold change as ≥2.5 for miRNA, and fold change as ≥1.9 for genes as thresholds to reduce false positives. If a gene/miRNA was expressed higher in the normal group compared to the tumor group, we defined that gene/miRNA as a down-regulated gene/miRNA, and vice versa.

Following this, we paired the miRNAs and mRNAs with different expression trends. For example, an up-regulated miRNA would be paired with a down-regulated mRNA. If such pairs could be found in 2 (or more) of the 3 miRNA target prediction databases, they were then added to their corresponding miRNA-regulated network. To further extend the coverage of our network, we incorporated the human protein reference database (HPRD) [47], which contains experimentally verified interaction data, into our miRNA-regulated PINs.

Finally, we used gene ontology (GO) enrichment analysis to explore the function of the miRNA-regulated PINs. Hypergeometric tests were used to determine if a GO term was enriched in a PIN. In this section, we excluded PINs with less than 5 proteins. We also excluded GO terms with levels of less than 5 to avoid non-specific GO terms. Since we tested multiple GO terms on each miRNA-regulated PIN, we adjusted the significance of the test with the FDR method developed by Benjamini et al.[48]. We used the adjusted p-value < 0.0001 as our threshold. A GO term would be excluded if its p-value was larger than 0.0001. Therefore, for each network, we ran the GO enrichment analysis, collected the calculated p-values, and adjusted these values using the methods described above.

3.5. ROC and GOBO Survival Analysis

After the PINs were constructed, we attempted to verify our results by literature search, ROC, and GOBO survival analysis [49]. To determine if the miRNAs we found could serve as classification markers for discriminating between normal and tumor samples, we applied ROC analysis on our miRNA array data (described in Section 2.1). ROC analysis is usually used to evaluate the efficiency of a classifier or a biological marker. R package ROCR [50] (version 1.0.4) was used to plot the ROC curve and calculate the area under curve (AUC). The standard error of AUC was then calculated as described in the work of Hanlye and McNeil [51]. The p-value of AUC was thus calculated with standard error obtained in the previous step. To further validate if the PINs we found were related to cancer, we used survival analysis implemented in GOBO [49] (available at, which provides a large amount of breast cancer gene expression profiles collected from public databases with clinical outcome data. In both ROC analysis and GOBO survival analysis, we considered our results significant when the p-value was smaller than 0.05.

4. Conclusions

Using integrative analysis of miRNA and mRNA expression profiles, we have identified not only breast cancer-related miRNAs and genes, but also putative roles for miRNAs in cancer as elucidated from miRNA-regulated PINs constructed in this work. Here, some previously known functions of miRNAs were again presented in our results, e.g., the relationship between the miRNAs, let-7c, miR-125a-5p, miR-125b-5p, and miR-21-5p, and breast cancer were demonstrated in this research. Furthermore, we have identified additional miRNAs and their related functions that have not been previously reported or discussed, providing valuable resources for further research in breast cancer.

Supplementary Information

Supplementary Information


This work was supported by the National Science Council of Taiwan (NSC 97-2311-B-002-010-MY3 and NSC 99-2621-B-002-005-MY3), National Taiwan University Cutting-Edge Steering Research Project (NTU-CESRP-102R7602C3) and the National Health Research Institutes, Taiwan (NHRI-EX98-9819PI).

Conflict of Interest

The authors declare no conflict of interest.


  1. Jemal, A.; Center, M.M.; DeSantis, C.; Ward, E.M. Global patterns of cancer incidence and mortality rates and trends. Cancer Epidemiol. Biomark. Prev 2010, 19, 1893–1907.
  2. Veronesi, U.; Boyle, P.; Goldhirsch, A.; Orecchia, R.; Viale, G. Breast cancer. Lancet 2005, 365, 1727–1741.
  3. Cho, E.; Spiegelman, D.; Hunter, D.J.; Chen, W.Y.; Stampfer, M.J.; Colditz, G.A.; Willett, W.C. Premenopausal fat intake and risk of breast cancer. J. Natl. Cancer Inst 2003, 95, 1079–1085.
  4. Huttenhofer, A.; Schattner, P.; Polacek, N. Non-coding RNAs: Hope or hype? Trends Genet 2005, 21, 289–297.
  5. Kim, V.N.; Han, J.; Siomi, M.C. Biogenesis of small RNAs in animals. Nat. Rev. Mol. Cell Biol 2009, 10, 126–139.
  6. Kwak, P.B.; Iwasaki, S.; Tomari, Y. The microRNA pathway and cancer. Cancer Sci 2010, 101, 2309–2315.
  7. Hutvagner, G.; Zamore, P.D. A microRNA in a multiple-turnover RNAi enzyme complex. Science 2002, 297, 2056–2060.
  8. Mourelatos, Z.; Dostie, J.; Paushkin, S.; Sharma, A.; Charroux, B.; Abel, L.; Rappsilber, J.; Mann, M.; Dreyfuss, G. miRNPs: A novel class of ribonucleoproteins containing numerous microRNAs. Genes Dev 2002, 16, 720–728.
  9. Nelson, P.T.; Hatzigeorgiou, A.G.; Mourelatos, Z. miRNP:mRNA association in polyribosomes in a human neuronal cell line. RNA 2004, 10, 387–394.
  10. Meister, G.; Landthaler, M.; Patkaniowska, A.; Dorsett, Y.; Teng, G.; Tuschl, T. Human Argonaute2 mediates RNA cleavage targeted by miRNAs and siRNAs. Mol. Cell 2004, 15, 185–197.
  11. Calin, G.A.; Dumitru, C.D.; Shimizu, M.; Bichi, R.; Zupo, S.; Noch, E.; Aldler, H.; Rattan, S.; Keating, M.; Rai, K.; et al. Frequent deletions and down-regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic leukemia. Proc. Natl. Acad. Sci. USA 2002, 99, 15524–15529.
  12. Zhang, B.; Pan, X.; Cobb, G.P.; Anderson, T.A. microRNAs as oncogenes and tumor suppressors. Dev. Biol 2007, 302, 1–12.
  13. Meng, F.; Henson, R.; Wehbe-Janek, H.; Ghoshal, K.; Jacob, S.T.; Patel, T. MicroRNA-21 regulates expression of the PTEN tumor suppressor gene in human hepatocellular cancer. Gastroenterology 2007, 133, 647–658.
  14. Rask, L.; Balslev, E.; Jorgensen, S.; Eriksen, J.; Flyger, H.; Moller, S.; Hogdall, E.; Litman, T.; Nielsen, B.S. High expression of miR-21 in tumor stroma correlates with increased cancer cell proliferation in human breast cancer. APMIS 2011, 119, 663–673.
  15. Zaman, M.S.; Shahryari, V.; Deng, G.; Thamminana, S.; Saini, S.; Majid, S.; Chang, I.; Hirata, H.; Ueno, K.; Yamamura, S.; et al. Up-regulation of microRNA-21 correlates with lower kidney cancer survival. PLoS One 2012, 7, e31060.
  16. Nadiminty, N.; Tummala, R.; Lou, W.; Zhu, Y.; Shi, X.B.; Zou, J.X.; Chen, H.; Zhang, J.; Chen, X.; Luo, J.; et al. MicroRNA let-7c is downregulated in prostate cancer and suppresses prostate cancer growth. PLoS One 2012, 7, e32832.
  17. Wang, X.F.; Shi, Z.M.; Wang, X.R.; Cao, L.; Wang, Y.Y.; Zhang, J.X.; Yin, Y.; Luo, H.; Kang, C.S.; Liu, N.; et al. MiR-181d acts as a tumor suppressor in glioma by targeting K-ras and Bcl-2. J. Cancer Res. Clin. Oncol 2012, 138, 573–584.
  18. Zhang, Y.; Yan, L.X.; Wu, Q.N.; Du, Z.M.; Chen, J.; Liao, D.Z.; Huang, M.Y.; Hou, J.H.; Wu, Q.L.; Zeng, M.S.; et al. miR-125b is methylated and functions as a tumor suppressor by regulating the ETS1 proto-oncogene in human invasive breast cancer. Cancer Res 2011, 71, 3552–3562.
  19. Sachdeva, M.; Mo, Y.Y. MicroRNA-145 suppresses cell invasion and metastasis by directly targeting mucin 1. Cancer Res 2010, 70, 378–387.
  20. Guo, X.; Wu, Y.; Hartley, R.S. MicroRNA-125a represses cell growth by targeting HuR in breast cancer. RNA Biol 2009, 6, 575–583.
  21. Jin, H.; Tuo, W.; Lian, H.; Liu, Q.; Zhu, X.Q.; Gao, H. Strategies to identify microRNA targets: New advances. N. Biotechnol 2010, 27, 734–738.
  22. Friedman, R.C.; Farh, K.K.; Burge, C.B.; Bartel, D.P. Most mammalian mRNAs are conserved targets of microRNAs. Genome Res 2009, 19, 92–105.
  23. Huang, J.C.; Babak, T.; Corson, T.W.; Chua, G.; Khan, S.; Gallie, B.L.; Hughes, T.R.; Blencowe, B.J.; Frey, B.J.; Morris, Q.D. Using expression profiling data to identify human microRNA targets. Nat. Methods 2007, 4, 1045–1049.
  24. Liu, B.; Li, J.; Tsykin, A.; Liu, L.; Gaur, A.B.; Goodall, G.J. Exploring complex miRNA-mRNA interactions with Bayesian networks by splitting-averaging strategy. BMC Bioinforma 2009, 10, 408.
  25. Tseng, C.W.; Lin, C.C.; Chen, C.N.; Huang, H.C.; Juan, H.F. Integrative network analysis reveals active microRNAs and their functions in gastric cancer. BMC Syst. Biol 2011, 5, 99.
  26. Farazi, T.A.; Horlings, H.M.; Ten Hoeve, J.J.; Mihailovic, A.; Halfwerk, H.; Morozov, P.; Brown, M.; Hafner, M.; Reyal, F.; van Kouwenhove, M.; et al. MicroRNA sequence and expression analysis in breast tumors by deep sequencing. Cancer Res 2011, 71, 4443–4453.
  27. Derfoul, A.; Juan, A.H.; Difilippantonio, M.J.; Palanisamy, N.; Ried, T.; Sartorelli, V. Decreased microRNA-214 levels in breast cancer cells coincides with increased cell proliferation, invasion and accumulation of the Polycomb Ezh2 methyltransferase. Carcinogenesis 2011, 32, 1607–1614.
  28. Heyn, H.; Engelmann, M.; Schreek, S.; Ahrens, P.; Lehmann, U.; Kreipe, H.; Schlegelberger, B.; Beger, C. MicroRNA miR-335 is crucial for the BRCA1 regulatory cascade in breast cancer development. Int. J. Cancer 2011, 129, 2797–2806.
  29. Sempere, L.F.; Christensen, M.; Silahtaroglu, A.; Bak, M.; Heath, C.V.; Schwartz, G.; Wells, W.; Kauppinen, S.; Cole, C.N. Altered MicroRNA expression confined to specific epithelial cell subpopulations in breast cancer. Cancer Res 2007, 67, 11612–11620.
  30. Zhu, X.M.; Wu, L.J.; Xu, J.; Yang, R.; Wu, F.S. Let-7c microRNA expression and clinical significance in hepatocellular carcinoma. J. Int. Med. Res 2011, 39, 2323–2329.
  31. Iyevleva, A.G.; Kuligina, E.; Mitiushkina, N.V.; Togo, A.V.; Miki, Y.; Imyanitov, E.N. High level of miR-21, miR-10b, and miR-31 expression in bilateral vs. unilateral breast carcinomas. Breast Cancer Res. Treat 2012, 131, 1049–1059.
  32. Rutnam, Z.J.; Yang, B.B. The involvement of microRNAs in malignant transformation. Histol. Histopathol 2012, 27, 1263–1270.
  33. Liu, H. MicroRNAs in breast cancer initiation and progression. Cell Mol. Life Sci 2012, 69, 3587–3599.
  34. Zhang, Z.J.; Ma, S.L. miRNAs in breast cancer tumorigenesis (Review). Oncol. Rep 2012, 27, 903–910.
  35. Mathivanan, S.; Periaswamy, B.; Gandhi, T.K.; Kandasamy, K.; Suresh, S.; Mohmood, R.; Ramachandra, Y.L.; Pandey, A. An evaluation of human protein-protein interaction data in the public domain. BMC Bioinforma 2006, 7, S19.
  36. Halbert, C.H.; Kumanyika, S.; Bowman, M.; Bellamy, S.L.; Briggs, V.; Brown, S.; Bryant, B.; Delmoor, E.; Johnson, J.C.; Purnell, J.; et al. Participation rates and representativeness of African Americans recruited to a health promotion program. Health Educ. Res 2010, 25, 6–13.
  37. Nicolas Diaz-Chico, B.; German Rodriguez, F.; Gonzalez, A.; Ramirez, R.; Bilbao, C.; Cabrera de Leon, A.; Aguirre Jaime, A.; Chirino, R.; Navarro, D.; Diaz-Chico, J.C. Androgens and androgen receptors in breast cancer. J. Steroid Biochem. Mol. Biol 2007, 105, 1–15.
  38. Sethi, N.; Kang, Y. Dysregulation of developmental pathways in bone metastasis. Bone 2011, 48, 16–22.
  39. Liao, S.J.; Zhou, Y.H.; Yuan, Y.; Li, D.; Wu, F.H.; Wang, Q.; Zhu, J.H.; Yan, B.; Wei, J.J.; Zhang, G.M.; et al. Triggering of Toll-like receptor 4 on metastatic breast cancer cells promotes alphavbeta3-mediated adhesion and invasive migration. Breast Cancer Res. Treat 2011, 133, 853–863.
  40. Jiang, P.; Enomoto, A.; Takahashi, M. Cell biology of the movement of breast cancer cells: Intracellular signalling and the actin cytoskeleton. Cancer Lett 2009, 284, 122–130.
  41. Garcia, D.M.; Baek, D.; Shin, C.; Bell, G.W.; Grimson, A.; Bartel, D.P. Weak seed-pairing stability and high target-site abundance decrease the proficiency of lsy-6 and other microRNAs. Nat. Struct. Mol. Biol 2011, 18, 1139–1146.
  42. Lewis, B.P.; Burge, C.B.; Bartel, D.P. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 2005, 120, 15–20.
  43. Lall, S.; Grun, D.; Krek, A.; Chen, K.; Wang, Y.L.; Dewey, C.N.; Sood, P.; Colombo, T.; Bray, N.; Macmenamin, P.; et al. A genome-wide map of conserved microRNA targets in C. elegans. Curr. Biol 2006, 16, 460–471.
  44. Krek, A.; Grun, D.; Poy, M.N.; Wolf, R.; Rosenberg, L.; Epstein, E.J.; MacMenamin, P.; da Piedade, I.; Gunsalus, K.C.; Stoffel, M.; et al. Combinatorial microRNA target predictions. Nat. Genet 2005, 37, 495–500.
  45. Betel, D.; Wilson, M.; Gabow, A.; Marks, D.S.; Sander, C. The resource: Targets and expression. Nucleic Acids Res 2008, 36, D149–D153.
  46. Tusher, V.G.; Tibshirani, R.; Chu, G. Significance analysis of microarrays applied to the ionizing radiation response. Proc. Natl. Acad. Sci. USA 2001, 98, 5116–5121.
  47. Keshava Prasad, T.S.; Goel, R.; Kandasamy, K.; Keerthikumar, S.; Kumar, S.; Mathivanan, S.; Telikicherla, D.; Raju, R.; Shafreen, B.; Venugopal, A.; et al. Human Protein Reference Database—2009 update. Nucleic Acids Res 2009, 37, D767–D772.
  48. Benjamini, Y.; Yekutieli, D. The control of the false discovery rate in multiple testing under dependency. Ann. Stat 2001, 29, 1165–1188.
  49. Ringner, M.; Fredlund, E.; Hakkinen, J.; Borg, A.; Staaf, J. GOBO: Gene expression-based outcome for breast cancer online. PLoS One 2011, 6, e17911.
  50. Sing, T.; Sander, O.; Beerenwinkel, N.; Lengauer, T. ROCR: Visualizing classifier performance in R. Bioinformatics 2005, 21, 3940–3941.
  51. Hanley, J.A.; McNeil, B.J. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982, 143, 29–36.
Ijms 14 11560f1 200
Figure 1. Analysis flow chart used in this work. After expression profiles and target prediction databases were fetched and preprocessed, they were subjected to the analysis process described here and in the “Experimental Section”.

Click here to enlarge figure

Figure 1. Analysis flow chart used in this work. After expression profiles and target prediction databases were fetched and preprocessed, they were subjected to the analysis process described here and in the “Experimental Section”.
Ijms 14 11560f1 1024
Ijms 14 11560f2 200
Figure 2. Receiver operating characteristic (ROC) curve of let-7c from our miRNA array dataset. For ROC curves of other miRNAs, see Figure S1.

Click here to enlarge figure

Figure 2. Receiver operating characteristic (ROC) curve of let-7c from our miRNA array dataset. For ROC curves of other miRNAs, see Figure S1.
Ijms 14 11560f2 1024
Ijms 14 11560f3 200
Figure 3. The let-7c-regulated protein interaction network (PIN). This is one of the 18 miRNA-regulated PINs constructed in this work. Figures of all other miRNA-regulated PINs are displayed in Supplementary Figures S2–S13.

Click here to enlarge figure

Figure 3. The let-7c-regulated protein interaction network (PIN). This is one of the 18 miRNA-regulated PINs constructed in this work. Figures of all other miRNA-regulated PINs are displayed in Supplementary Figures S2–S13.
Ijms 14 11560f3 1024
Ijms 14 11560f4 200
Figure 4. Validation of let-7c result. Gene expression-based outcome for breast cancer online (GOBO) survival analysis of let-7c-regulated PIN members marked with the following functions: (A) Microtubule cytoskeleton; (B) Negative regulation of programmed cell death; and (C) negative regulation of cell death. Red: samples with high expression of selected gene set (PIN members); grey: samples with low expression of selected gene set (PIN members).

Click here to enlarge figure

Figure 4. Validation of let-7c result. Gene expression-based outcome for breast cancer online (GOBO) survival analysis of let-7c-regulated PIN members marked with the following functions: (A) Microtubule cytoskeleton; (B) Negative regulation of programmed cell death; and (C) negative regulation of cell death. Red: samples with high expression of selected gene set (PIN members); grey: samples with low expression of selected gene set (PIN members).
Ijms 14 11560f4 1024
Table 1. Selected enriched functions of let-7c. Member genes of the let-7c-regulated network annotated with corresponding enriched functions are listed.

Click here to display table

Table 1. Selected enriched functions of let-7c. Member genes of the let-7c-regulated network annotated with corresponding enriched functions are listed.
GO termGenesAdj. p-value
GO:0043067, Regulation of programmed cell deathRRM2B, ACVR1, TP53, RASA1, TGFBR1, PSMA3, BIRC5, ACTN2, HOXA13, IRS2, FASTK, VAV1, PSMB6, BCL2, CDK1, HDAC1, SOX10, TIA1, AKT1, AURKB3.43 × 10−8
GO:0043069, Negative regulation of programmed cell deathRRM2B, ACVR1, TP53, RASA1, TGFBR1, BIRC5, IRS2, BCL2, CDK1, HDAC1, SOX10, AKT1, AURKB6.58 × 10−6
GO:0060548, Negative regulation of cell deathRRM2B, ACVR1, TP53, RASA1, TGFBR1, BIRC5, IRS2, BCL2, CDK1, HDAC1, SOX10, AKT1, AURKB8.92 × 10−6
GO:0015630, Microtubule cytoskeletonINCENP, SNTB2, SEPT1, TACC1, BIRC5, RACGAP1, PIN4, CDCA8, CDK1, PHF1, AKT1, AURKB, NINL, CCDC85B5.15 × 10−5
Int. J. Mol. Sci. EISSN 1422-0067 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert