Purpose: Recent studies have suggested that microRNA biomarkers could be useful for stratifying lung cancer subtypes, but microRNA signatures varied between different populations. Squamous cell carcinoma (SCC) is one major subtype of lung cancer that urgently needs biomarkers to aid patient management. Here, we undertook the first comprehensive investigation on microRNA in Chinese SCC patients.

Experimental Design: MicroRNA expression was measured in cancerous and noncancerous tissue pairs strictly collected from Chinese SCC patients (stages I–III), who had not been treated with chemotherapy or radiotherapy prior to surgery. The molecular targets of proposed microRNA were further examined.

Results: We identified a 5-microRNA classifier (hsa-miR-210, hsa-miR-182, hsa-miR-486-5p, hsa-miR-30a, and hsa-miR-140-3p) that could distinguish SCC from normal lung tissues. The classifier had an accuracy of 94.1% in a training cohort (34 patients) and 96.2% in a test cohort (26 patients). We also showed that high expression of hsa-miR-31 was associated with poor survival in these 60 SCC patients by Kaplan–Meier analysis (P = 0.007), by univariate Cox analysis (P = 0.011), and by multivariate Cox analysis (P = 0.011). This association was independently validated in a separate cohort of 88 SCC patients (P = 0.008, 0.011, and 0.003 in Kaplan–Meier analysis, univariate Cox analysis, and multivariate Cox analysis, respectively). We then determined that the tumor suppressor DICER1 is a target of hsa-miR-31. Expression of hsa-miR-31 in a human lung cancer cell line repressed DICER1 activity but not PPP2R2A or LATS2.

Conclusions: Our results identified a new diagnostic microRNA classifier for SCC among Chinese patients and a new prognostic biomarker, hsa-miR-31. Clin Cancer Res; 17(21); 6802–11. ©2011 AACR.

Translational Relevance

The poor outcome of patients with squamous cell carcinoma (SCC) is exacerbated by the lack of early diagnosis markers and the lack of prognostic indicators. The 5-microRNA classifier distinguishing SCC from normal lung tissues holds promise for early diagnosis of SCC among Chinese. The classifier may potentially improve the diagnostic value of transbronchial lung biopsy. hsa-miR-31 was identified conclusively as a prognostic factor and may be of use for the clinical management of SCC.

Worldwide, there are more than 1,600,000 new cases of lung cancer and more than 1,370,000 attributable deaths each year, ranking it as the leading cause of cancer mortality. In China, lung cancer is the most frequent of the malignant tumors, initiating 520,000 new cases and causing 450,000 deaths per year; furthermore, the age-standardized incidence rate and mortality rate adjusted for the world standard population are both higher than the worldwide average levels (1). Non–small cell lung cancers (NSCLC) are the major types of lung cancer found worldwide (∼80%). Among NSCLCs, squamous cell carcinoma (SCC, 40%) is one of the main pathologic subtypes in China. SCC patients with clinical stage IA disease have a 5-year survival rate of about 60%, whereas the 5-year survival rate for clinical stage II to IV disease ranges from 40% to less than 5% (2). The poor outcome of SCC patients could be explained, in part, by the lack of early diagnosis markers and the lack of prognostic indicators. Therefore, there is an urgent need to develop biomarkers for diagnosis and prognosis, to benefit Chinese SCC patients. Transbronchial lung biopsy (TBLB) has been explored as a method for early diagnosis of questionable lung infiltrations or densities; however, the diagnostic yield of TBLB is lower than that of open lung biopsy because the small amount of lung tissues obtained in biopsy. New biomarkers with high diagnostic potential to distinguish SCC cancerous from normal lung tissue should aid the acute diagnostic yield. Furthermore, novel prognostic biomarkers will ultimately allow SCC patients to receive personalized therapies.

MicroRNAs are a species of small (19–22 nucleotides) noncoding single-stranded RNA molecules that through partial sequence homology may interact with the 3′-untranslated region (3′-UTR) of target mRNA molecules (3, 4). Different microRNAs may function as tumor suppressors or oncogenes, and the deregulation of their expression is associated with the initiation and progression of cancers via the activation and/or repression of controlling pathways (5). Several recent studies have shown that selected microRNAs contributing to NSCLC progression can also be used to estimate prognosis (6–11). Although microRNA profiling of NSCLCs has contributed to our understanding of the biology of these cancers, the microRNA signatures identified in NSCLC (including adenocarcinoma and SCC) were not consistent among different clinical studies. The variations were likely caused by factors such as the differences among the samples and methodologies used, including the ethnicity of patient groups, the pathologic subtypes collected, the sample collection methods, the technology platforms [microarray or quantitative reverse transcription PCR (qRT-PCR)], and the bioinformatics analysis used. The importance of carefully designed and appropriately controlled studies is illustrated by the well-established, although very different, biomarker profiles used currently to assess breast cancer prognosis with Agendia's MammaPrint and Genomic Health's Oncotype DX genomic assays. The MammaPrint assay comprises 70 genes that were found to be related to distant recurrence of breast cancer, whereas Oncotype DX evaluates the activity of 21 genes that indicate the risk of breast cancer recurrence. Only one gene is common between the 2 tests, yet both have been used effectively by thousands of doctors to help guide treatment of more than 200,000 patients (http://www.oncotypedx.com/en-US/Breast/HealthcareProfessional/Overview.aspx, http://investor.genomichealth.com/releasedetail.cfm?ReleaseID=386285, and http://www.mammaprint.co.uk/MammaPrint_Patient_Brochure_EU_lowres.pdf). To identify promising microRNA biomarkers for lung cancer subtypes, it is necessary that the study is critically designed and that an established and proven methodology is employed. We also ensured that only a specific pathologic cancer subtype (SSC) in one ethnic group (Han Chinese) was examined, and we carefully chose the best of collected patient samples to limit outlier contamination. The microarray platform and data analysis methods were also well established.

We conducted a genome-wide microRNA study (677 human microRNA probes) to ensure a broad profiling among cellular microRNAs. Principal component analysis (PCA) and support vector machine (SVM) analysis tools were then used to establish a powerful microRNA classifier that distinguished SCC from normal tissue: a minimal 5-element classifier (hsa-miR-210, hsa-miR-182, hsa-miR-486-5p, hsa-miR-30a, and hsa-miR-140-3p) was found. The patients chosen for our study had comprehensive follow-up data that allowed prognostic analysis, which then revealed that the high expression of hsa-miR-31 was associated with poor survival from SCC. This marker was confirmed by Kaplan–Meier analysis, univariate Cox analysis, and multivariate Cox analysis. Finally, we showed a molecular mechanism by which hsa-miR-31 might contribute to poor patient prognosis. hsa-miR-31 targets the tumor suppressor DICER1, resulting in the repression of its expression.

Patients and samples

The patients examined in this study underwent surgery at the Cancer Institute and Hospital, Chinese Academy of Medical Sciences, between 2000 and 2002. All patients (Supplementary Table S1) had not been previously treated by chemotherapy and radiotherapy when undergoing surgery and provided informed consent to participate in the study. The study was approved by the medical ethics committee of the Cancer Institute and Hospital, Chinese Academy of Medical Sciences.

Samples were snap frozen in liquid nitrogen immediately after resection and stored (minimum of 5 years) at −80°C until the extraction of RNA. Peripheral portions of the resected lung samples were paraffin embedded, sectioned, and hematoxylin and eosin stained by routine methods. The tumor cell concentrations were evaluated, and the tumor histology was independently confirmed by 2 pathologists (S. Shi and X. Feng). Follow-up information was extracted from the follow-up registry of the Cancer Institute and Hospital, Chinese Academy of Medical Sciences. For all the samples, clinicopathologic information [smoking, age, gender, pathologic subtype, tumor node metastasis (TNM) classification, tumor stage, lymph node stage, differentiation status, adjuvant therapy after surgery, and the duration of survival after surgery] was available.

We initially profiled the microRNA expression in a training cohort of 34 SCC patients for classifier establishment, using paired cancerous and adjacent normal samples, and then examined an additional 26 SCC patients as a test cohort. The microRNA expression in this total of 60 SCC patients was used for prognostic analysis of their related survival data, and a further cohort of 88 SCC patients was used as an independent validation of the prognostic analysis.

MicroRNA microarray analysis

The microRNA analyses were conducted on a microarray using the CapitalBio platform (CapitalBio Corp.) that has been used in numerous previous studies (12, 13). Total RNA was extracted with TRIzol reagent (Invitrogen), and the low-molecular-weight RNA was isolated by a polyethylene glycol (PEG) solution precipitation method (14). The low-molecular-weight RNA was labeled by the T4 RNA ligase labeling method described by Thomson and colleagues (15). In brief, 4 μg of low-molecular-weight RNA isolated by a PEG solution precipitation method was labeled with 500 ng of 5′-phophate-cydidyl-urdyl-cy3-3′, using the T4 RNA ligase. The labeled RNA was hybridized with aldehyde-modified slide containing 924 mature mammalian microRNA probes (including 677 human microRNA sequences), with each probe present in triplicate spots. A 3-dimensional tilting agitator BioMixer II (CapitalBio) was used for overnight hybridization to increase probe signal intensity and improve signal uniformity. An automated SlideWasher-8 (CapitalBio) was used to wash and dry the hybridized slides to reduce the dye blemishes that occur frequently during manual posthybridization slide washing. Arrays were scanned with a LuxScan 10K-A laser confocal scanner, and the images obtained were then analyzed with LuxScan 3.0 software (both from CapitalBio).

Bioinformatics analysis

For all samples, the average background values were subtracted from each of the replicate microRNA spots. Faint spots were filtered out if the expression signal was less than 1,500. A space- and intensity-dependent median-center normalization was done with Bionconductor. Differentially expressed microRNAs were identified by significance analysis of microarrays (SAM; (http://www-stat.stanford.edu/∼tibs/SAM/index.html; ref. 16). The sufficient but minimal microRNA marker set to classify the cancerous and adjacent normal tissues was analyzed by the PCA and SVM methods (12). The 632 Bootstrap method was used to estimate the accuracy of each predicted model for the original training set by the use of random resampling with replacement over 1,000 independent analyses (17). The accuracy was calculated by the formula, ${\rm acc}_{{\rm boot}} = {\frac{1}{n}}\sum\nolimits_{i = 1}^n {\left({0.368 \times {\rm acc}_{{\rm train}}^i + 0.632 \times {\rm acc}_{{\rm test}}^i} \right)}$⁠, where n is the number of repeats, accitrain is the ith experiment train accuracy, and $acc_{{\rm test}}^i$ is the ith experiment test accuracy. The possible gene functions of the most significant predicted microRNA targets were analyzed by 4 publicly available algorithms, miRBase (http://microrna.sanger.ac.uk/sequences/; ref. 18), MiRanda (http://www.microrna.org; ref. 19), TargetScan (http://www.targetscan.org/; ref. (20), and PicTar (http://pictar.bio.nyu.edu/; ref. 21). To reduce the number of false-positive results, only putative target genes predicted by at least 3 of the programs were accepted. All microRNA expression data have been submitted to the Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo) with the series accession number GSE 15008.

Survival analysis

Patient survival curves were estimated by the Kaplan–Meier method. The median microRNA intensity value of the initial training cohort was used as the cutoff point in Kaplan–Meier survival analysis, and patients were categorized into groups with high (above and equal median) or low (below median) expression. The joint effect of covariables was examined by the Cox proportional hazard regression model.

qRT-PCR analysis

Total cellular RNAs were also subjected to qRT-PCR with microRNA specific primers for determination of microRNA expression. Total RNA extraction, reverse transcription, and quantitative PCR were done according to the procedures described previously (12, 22). The primers for hsa-miR-31 are as follows: RT primer, 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACAGCTAT-3′; amplification primer, 5′-AGGCAAGATGCTGGCATAGCT-3′ and 5′-CAAGGCAAGATGCTGGCATA-3′. For the quantification of DICER1 mRNA, total cellular RNA was reverse transcribed with oligo(dT) primers. Amplification of DICER1 was done with primers 5′-CACTCCCATCAACATACCAGTAGA-3′ and 5′-ACGAAGCATCAAGTTCAGAATCAT3′. The highly conserved and universally expressed small nuclear RNA U6 and β-actin genes were used as endogenous qRT-PCR controls for hsa-miR-31 and DICER1, respectively.

Cell culture and transient transfection

Human lung squamous cell carcinoma SK-MES-1 cells (2 × 105 per well) were plated into a 6-well plate 1 day before transfection. hsa-miR-31 or control microRNA (Dharmacon) at a concentration of 100 nmol/L was transfected into cells with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Cell lysates were prepared for Western blot assay at 48 hours posttransfection and luciferase reporter assay at 24 hours posttransfection.

Western blot analysis

SK-MES-1 cell lysates (20 μg) were separated by 10% SDS-PAGE and electroblotted onto polyvinylidene fluoride (PVDF) membrane (Pall). The PVDF membrane was blocked with TBS Tween-20 buffer containing 5% milk powder, and reaction with primary antibody was carried out at 4°C overnight with gentle shaking. After washing, secondary antibody conjugated with horseradish peroxidase (HRP) was added, and the reaction was carried out for 1 hour at room temperature. SuperSignal West Pico Chemiluminescent Substrate (Pierce) was used for the detection according to the manufacturer's instructions. Western blot results were scanned and quantified by NIH ImageJ software (http://rsbweb.nih.gov/ij/). Protein expression levels were normalized by the expression levels of β-tubulin.

Luciferase reporter assay

The 3′-UTRs of predicted microRNA target genes were amplified by PCR and cloned into the XbaI site of pRL-SV40 vector (Promega). Primers for the amplification of (i) DICER1 are 5′-GAGTGGAATGAATTGAAGGCAGAA-3′ and 5′-ATCACGCTGTCTCAACGTCTAATG-3′, (ii) PPP2R2A 3′-UTR are 5′-TTTCAAGACAAAGTGAATTAGGGTTGG-3′ and 5′-ACACGGCGAGGAGTGATTACAGC-3′, and (iii) LATS2 3′-UTR are 5′-GCTTTCAATAGGCTTTTCAGGACC-3′ and 5′-CATACGGACTACAGAAACGGACAT-3′. The mutation of DICER1 3′-UTR sequence was undertaken with the QuikChange Site–Directed Mutagenesis Kit (Stratagene) with primers 5′-AGTCTTGCATAAAAAAGGGTTCTACGCTTAAAAGTGAAACCTTCATGGAT3′ and 5′-ATCCATGAAGGTTTCACTTTTAAGCGTAGAACCCTTTTTTATGCAAGACT-3′. SK-MES-1 cells were transfected with a mixture of pGL3-control, pRL-SV40-3′-UTR, and miR-31 or control microRNA (Dharmacon). Cells were lysed at 24 hours posttransfection. Firefly and Renilla luciferase activities were assayed with the Dual-Glo Luciferase Assay System (Promega) according to the manufacturer's instructions. Data were normalized to the control microRNA.

Immunohistochemical assay

Anti-DICER1 (catalog #ab82539) was purchased from Abcam. HRP-labeled goat anti-mouse IgG was purchased from Abgent. Immunohistochemistry was carried out on 4-μm, formalin-fixed, paraffin-embedded (FFPE) sections with anti-DICER1 as described (23). Briefly, the slides were deparaffinized, rehydrated, and dripped in 3% hydrogen peroxide solution for 10 minutes. The slides were incubated with anti-DICER1 (1:50 dilution) at 4°C for 8 hours. After washing with PBS, the slides were then incubated with Polymer Helper for 20 minutes, followed by polyperoxidase–anti-mouse IgG for 30 minutes at room temperature. Development of the slides was carried out using 3,3′-diaminobenzidine solution. Counterstaining was carried out with hematoxylin. Negative control was obtained by substituting the primary antibody with PBS. Immunohistochemical results were evaluated by 2 pathologists (S. Shi and X. Feng) without previous knowledge of patient information. The expression was assigned to one of the following categories according to the percentage of positive cells: 0 (0%–5%), + (6%–25%), ++ (26%–50%), and +++ (51%–100%). The expression was considered negative when the score was 0 and positive when the score was + or greater. An unpaired Student t test was used for analysis of data between groups. Values of P < 0.05 were considered statistically significant.

A 5-microRNA signature can efficiently distinguish malignant SCC tissues from adjacent normal tissues in Chinese

We first investigated whether the microRNA expression could identify malignant SCC tissues in Chinese patients by examining primary lung cancers and their corresponding adjacent normal lung tissues collected a minimum distance of 5 cm from the tumor in 34 patients (training cohort). Twenty-two microRNAs (Supplementary Table S2) were found differentially expressed between cancer and normal tissues, which were further analyzed by the PCA–SVM strategy, and a minimum of 5-microRNA classifier set (hsa-miR-210, hsa-miR-182, hsa-miR-486-5p, hsa-miR-30a, and hsa-miR-140-3p), with the highest distinguishing values, was identified. The analysis of the original training group using this classifier had a predictive accuracy of 94.1%. The classifier was subsequently validated with an independent test cohort comprising another 26 SCC patients and displayed an accuracy of 96.2% with this group (Fig. 1). In the SCC tissues, 2 of the microRNAs (hsa-miR-210 and hsa-miR-182) were upregulated and 3 microRNAs (hsa-miR-486-5p, hsa-miR-30a, and hsa-miR-140-3p) were downregulated. These analyses indicated that a classifier with as few as 5-microRNA markers could efficiently distinguish malignant SCC tissues from normal tissues in Chinese patients.

Figure 1.

A 5-microRNA classifier could distinguish malignant SCC lesions from normal tissues. The PCA–SVM strategy was used to construct a classifier, which was used to classify the cancerous tissue and adjacent normal tissue. Red dots represent the cancerous tissue and blue crosses represent the adjacent normal tissue. The classifier had a predictive accuracy of 94.1% in the original training cohort (34 SCC patients; A) and 96.2% in the independent test cohort (another 26 SCC patients; B).

Figure 1.

A 5-microRNA classifier could distinguish malignant SCC lesions from normal tissues. The PCA–SVM strategy was used to construct a classifier, which was used to classify the cancerous tissue and adjacent normal tissue. Red dots represent the cancerous tissue and blue crosses represent the adjacent normal tissue. The classifier had a predictive accuracy of 94.1% in the original training cohort (34 SCC patients; A) and 96.2% in the independent test cohort (another 26 SCC patients; B).

Close modal

High expression of hsa-miR-31 correlates strongly with low survival in Chinese SCC patients

We then examined the 22 differentially expressed microRNAs in 60 SCC patients (clinical characteristics listed in Table 1 and Supplementary Table S1) using Kaplan–Meier survival analysis for potential correlations between microRNA expression and survival prognosis. Kaplan–Meier analysis showed that a high expression of hsa-miR-31 was associated with poor survival in SCC patients (P = 0.007; log-rank test, 60 SCC patient training cohort; Fig. 2A). Subsequently, univariate Cox analysis with hsa-miR-31 and clinicopathologic factors (smoking, age, gender, differentiation, TNM classification, and adjuvant therapy after surgery) revealed that the hsa-miR-31 expression level had prognostic significance in the training cohort (P = 0.011; risk ratio: 2.827; 95% CI, 1.274–6.275; Table 2). Furthermore, multivariate Cox proportional hazard regression analysis of each of these parameters indicated that high hsa-miR-31 expression was a significantly unfavorable prognostic factor independent of all other clinicopathologic factors (P = 0.011; risk ratio 3.535; 95% CI, 1.340–9.328; Table 2). These results showed that high levels of hsa-miR-31 expression correlated strongly with poor survival of Chinese SCC patients.

Figure 2.

Kaplan–Meier survival curves for SCC patients. High levels of hsa-miR-31 expression correlated with a lower survival rate in the training cohort (60 SCC patients; A), whereas low expression correlated with a high survival rate in an independent test cohort (88 SCC patients; B). The log-rank test P value was 0.007 for the training cohort and 0.008 for the test cohort.

Figure 2.

Kaplan–Meier survival curves for SCC patients. High levels of hsa-miR-31 expression correlated with a lower survival rate in the training cohort (60 SCC patients; A), whereas low expression correlated with a high survival rate in an independent test cohort (88 SCC patients; B). The log-rank test P value was 0.007 for the training cohort and 0.008 for the test cohort.

Close modal
Table 1.

Clinical characteristics of the training and test cohorts for prognostic analysis in SCC patients

CharacteristicTraining cohort (n = 60)Test cohort (n =88)
Age, y 
 Range 17–77 42–81 
 Median 59.0 62.5 
Sex, n (%) 
 Male 50 (83) 79 (90) 
 Female 10 (17) 9 (10) 
Smoking, n (%) 
 Smoking 49 (82) 60 (68) 
 Nonsmoking 11 (18) 28 (32) 
Stage (TNM) classification, n (%) 
 I 21 (35) 27 (31) 
 II 17 (28) 31 (35) 
 III 22 (37) 30 (34) 
Differentiation, n (%) 
 Well 5 (8) 4 (5) 
 Moderate 33 (55) 44 (50) 
 Poor 22 (37) 40 (45) 
CharacteristicTraining cohort (n = 60)Test cohort (n =88)
Age, y 
 Range 17–77 42–81 
 Median 59.0 62.5 
Sex, n (%) 
 Male 50 (83) 79 (90) 
 Female 10 (17) 9 (10) 
Smoking, n (%) 
 Smoking 49 (82) 60 (68) 
 Nonsmoking 11 (18) 28 (32) 
Stage (TNM) classification, n (%) 
 I 21 (35) 27 (31) 
 II 17 (28) 31 (35) 
 III 22 (37) 30 (34) 
Differentiation, n (%) 
 Well 5 (8) 4 (5) 
 Moderate 33 (55) 44 (50) 
 Poor 22 (37) 40 (45) 
Table 2.

Postoperative survival of SCC patients in relation to clinicopathologic characteristics and microRNA expression analyzed by the Cox proportional hazard regression model in a training cohort of 60 cases

 Univariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Smoking Yes/no 1.014 (0.385–2.668) 0.978 1.859 (0.581–5.951) 0.515 
Age ≥59 1.111 (0.528–2.338) 0.781 1.149 (0.505–2.617) 0.740 
Sex M/F 1.265 (0.481–3.331) 0.634 1.997 (0.637–6.256) 0.235 
TNM I/II/III 1.794 (1.131–2.845) 0.013 1.325 (0.775–2.267) 0.304 
Differentiation Well/moderate/poor 1.283 (0.656–2.508) 0.466 2.053 (0.882–5.130) 0.124 
Adjuvant therapy Yes/no 0.833 (0.374–1.855) 0.655 1.163 (0.459–2.950) 0.750 
hsa-miRNA-31 High/low 2.827 (1.274–6.275) 0.011 3.535 (1.340–9.328) 0.011 
 Univariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Smoking Yes/no 1.014 (0.385–2.668) 0.978 1.859 (0.581–5.951) 0.515 
Age ≥59 1.111 (0.528–2.338) 0.781 1.149 (0.505–2.617) 0.740 
Sex M/F 1.265 (0.481–3.331) 0.634 1.997 (0.637–6.256) 0.235 
TNM I/II/III 1.794 (1.131–2.845) 0.013 1.325 (0.775–2.267) 0.304 
Differentiation Well/moderate/poor 1.283 (0.656–2.508) 0.466 2.053 (0.882–5.130) 0.124 
Adjuvant therapy Yes/no 0.833 (0.374–1.855) 0.655 1.163 (0.459–2.950) 0.750 
hsa-miRNA-31 High/low 2.827 (1.274–6.275) 0.011 3.535 (1.340–9.328) 0.011 

Validation of the survival correlation of hsa-miR-31 with an independent cohort of Chinese SCC patients

The correlation of hsa-miR-31 with the prognosis of SCC was examined further with an independent set of 88 SCC patients. The hsa-miR-31 expression was determined by qRT-PCR. Kaplan–Meier survival analysis confirmed a significantly decreased survival of patients with high hsa-miR-31 expression (P = 0.008; log-rank test, 88 SCC patient test cohort; Fig. 2B). Both univariate (P = 0.011; risk ratio, 2.396; 95% CI, 1.218–4.715) and multivariate (P = 0.003; risk ratio, 3.071; 95% CI, 1.451–6.500) Cox proportional hazard regression analyses suggested that high hsa-miR-31 expression was an independent predictor of poor prognosis for SCC in Chinese (Table 3).

Table 3.

Postoperative survival of SCC patients in relation to clinicopathologic characteristics and microRNA expression analyzed by the Cox proportional hazard regression model in a test cohort of 88 cases

 Univariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Smoking Yes/no 1.989 (0.908–4.355) 0.086 2.246 (0.983–5.130) 0.055 
Age ≥60 1.169 (0.606–2.254) 0.639 2.211 (1.020–4.793) 0.044 
Sex M/F 0.969 (0.343–2.737) 0.953 0.682 (0.211–2.209) 0.524 
TNM I/II/III 1.263 (0.839–1.899) 0.263 1.018 (0.662–1.566) 0.936 
Differentiation Well/moderate/poor 2.543 (1.354–4.779) 0.004 2.408 (1.237–4.687) 0.010 
Adjuvant therapy Yes/no 3.060 (1.342–6.976) 0.008 2.897 (1.181–7.105) 0.020 
hsa-miRNA-31 High/low 2.396 (1.218–4.715) 0.011 3.071 (1.451–6.500) 0.003 
 Univariate analysisMultivariate analysis
HR (95% CI)PHR (95% CI)P
Smoking Yes/no 1.989 (0.908–4.355) 0.086 2.246 (0.983–5.130) 0.055 
Age ≥60 1.169 (0.606–2.254) 0.639 2.211 (1.020–4.793) 0.044 
Sex M/F 0.969 (0.343–2.737) 0.953 0.682 (0.211–2.209) 0.524 
TNM I/II/III 1.263 (0.839–1.899) 0.263 1.018 (0.662–1.566) 0.936 
Differentiation Well/moderate/poor 2.543 (1.354–4.779) 0.004 2.408 (1.237–4.687) 0.010 
Adjuvant therapy Yes/no 3.060 (1.342–6.976) 0.008 2.897 (1.181–7.105) 0.020 
hsa-miRNA-31 High/low 2.396 (1.218–4.715) 0.011 3.071 (1.451–6.500) 0.003 

hsa-miR-31 targets DICER1 but not PPP2R2A and LATS2

We finally investigated the potential molecular targets of hsa-miR-31 by analysis of the target genes predicted by 4 algorithms (miRBase, MiRanda, TargetScan, and PicTar). We selected DICER1, PPP2R2A, and LATS2 for further investigation. The duplexes predicted to form between hsa-miR-31 and the 3 target genes are depicted in Figure 3A. Immunoblot analysis of DICER1, PPP2R2A, and LATS2 proteins showed that only DICER1 protein was reduced in SK-MES-1 cells transiently transfected with hsa-miR-31 (Fig. 3B). We carried out a dual luciferase reporter assay to confirm whether hsa-miR-31 target DICER1 mRNA. A significant decrease in luciferase activity was observed when we cotransfected hsa-miR-31 and luciferase reporter gene linked to the wild-type 3′-UTR of DICER1, whereas specific mutation of the hsa-miR-31 binding sequence in DICER1 3′-UTR resulted in only slightly downregulation of the reporter gene (Fig. 3C). We also constructed luciferase reporters linked to the 3′-UTR of PPP2R2A and of LATS2, but no significant effects on these reporter genes were observed in the presence of cotransfected hsa-miR-31 (data not shown). The immunoblot and reporter assays independently confirmed that DICER1 is a target of hsa-miR-31 in human cells. Interestingly, quantification of DICER1 mRNA indicated a certain downregulation of the level of DICER1 mRNA expression in SK-MES-1 cells transfected with hsa-miR-31 as compared with control microRNA, even though the P value is not significant, suggesting that hsa-miR-31 may act mainly by inhibiting the translation of DICER1 (Fig. 3D).

Figure 3.

hsa-miR-31 targets DICER1 and not PPP2R2A or LATS2. A, predicted duplex formation between hsa-miR-31 and human DICER1, PPP2R2A, and LATS2 3′-UTR. Arrows indicate the mutation of the hsa-miR-31 complementary sites in DICER1 3′-UTR. B, immunoblots of DICER1, PPP2R2A, and LATS2 proteins in SK-MES-1 cells transiently transfected with hsa-miR-31 and control microRNA, respectively. C, luciferase reporter assays of the interaction between hsa-miR-31 and the DICER1 3′-UTR. Assays were carried out by cotransfection of hsa-miR-31 with a luciferase reporter gene linked to the 3′-UTR of DICER1, containing either wild-type (WT) or mutated (mut) hsa-miR-31 complementary sites. D, qRT-PCR of DICER1 mRNA in SK-MES-1 cells transfected with hsa-miR-31 and control microRNA, respectively. A slight downregulation of the level of DICER1 mRNA expression can be seen, although the P value was not significant. E, the expression of DICER1 in tumor tissues (T) was examined by immunohistochemistry. These sections were also stained by hematoxylin and eosin (H&E); 100-fold and 200-fold indicate the magnification times. F, the correlation between hsa-miR-31 level and DICER1 expression was analyzed by the χ2 test, n = 46, P = 0.036.

Figure 3.

hsa-miR-31 targets DICER1 and not PPP2R2A or LATS2. A, predicted duplex formation between hsa-miR-31 and human DICER1, PPP2R2A, and LATS2 3′-UTR. Arrows indicate the mutation of the hsa-miR-31 complementary sites in DICER1 3′-UTR. B, immunoblots of DICER1, PPP2R2A, and LATS2 proteins in SK-MES-1 cells transiently transfected with hsa-miR-31 and control microRNA, respectively. C, luciferase reporter assays of the interaction between hsa-miR-31 and the DICER1 3′-UTR. Assays were carried out by cotransfection of hsa-miR-31 with a luciferase reporter gene linked to the 3′-UTR of DICER1, containing either wild-type (WT) or mutated (mut) hsa-miR-31 complementary sites. D, qRT-PCR of DICER1 mRNA in SK-MES-1 cells transfected with hsa-miR-31 and control microRNA, respectively. A slight downregulation of the level of DICER1 mRNA expression can be seen, although the P value was not significant. E, the expression of DICER1 in tumor tissues (T) was examined by immunohistochemistry. These sections were also stained by hematoxylin and eosin (H&E); 100-fold and 200-fold indicate the magnification times. F, the correlation between hsa-miR-31 level and DICER1 expression was analyzed by the χ2 test, n = 46, P = 0.036.

Close modal

Expression of DICER1 is inversely correlated with hsa-miR-31 level in SCC tumor tissues

To further determine the relationship between DICER1 and hsa-miR-31 in SCC tumor tissues, we carried out the immunohistochemical assay in 46 SCC samples for which the corresponding FFPE tumor tissues were available. We found that the level of DICER1 was markedly lower in SCC tumor tissues with high miRNA expression. The DICER1 expression was scored negative in 83% (20 of 24) tumor tissues with high hsa-miR-31 levels, whereas only 58% (12 of 22) tumor tissues with low hsa-miR-31 level were negative for DICER1 (P = 0.036; Fig. 3E and F). These results indicated that the expression of DICER1 is inversely correlated with hsa-miR-31 level in SCC tumor tissues.

The identification and validation of novel biomarkers for different cancers comprise a significant area of practical cancer research. Multiplex molecular diagnostic and prognostic tests for cancers have been intensively investigated during the past 10 years, and now including genomic DNA, mRNAs, microRNAs, DNA methylation, and proteins. Two demonstrable analytic successes are Agendia's MammaPrint and Genomic Health's Oncotype DX genomic assays, which are both used by thousands of doctors in clinical practice.

Several recent clinical studies have proposed sets of differentially expressed microRNAs as promising markers for monitoring the development of lung cancer and its prognosis, in particular for NSCLC (6–11). An important consideration is how to achieve a reliable and universal signature, considering that results from different studies that identified biomarker species often varied. The reason for this occurrence is complex but likely arises from the differences in patient ethnicities, pathologic sample subtypes, sample collection methods, technology platforms, bioinformatics approaches, as well as the different experience and expertise of the investigators with each of the methodologic variables.

We, therefore, took particular care in the design of this study to ensure the identification of reliable biomarkers and used only established methods that reveal valuable information. We assayed only Chinese SCC patients to minimize ethnicity and pathologic subtype effects and to amplify the molecular homogeneity of tumor specimens. The patients were strictly selected from individuals who had not been previously treated with chemotherapy or radiotherapy when undergoing surgery to avoid therapy influences. A microarray containing 677 mature human microRNA probes was used to determine expression levels. Although this microarray does not represent the latest newly discovered microRNAs (The miRBase database version 16 contains 1,212 mature human microRNA; http://www.mirbase.org/), the study still revealed new and important information about microRNA useful for the management of SCC. The MicroArray Quality Control Consortium (MAQC I) study previously reported the high reproducibility of expression analysis results produced by the microarray platform our group used here (24). More recently, the MAQC II study elucidated that the execution of a gene expression classifier depended largely on the proficiency of the researchers and on the different methods used (25). The PCA–SVM strategies used here for lung cancer microRNA profiling analysis are well established in our group and were used previously in a study of esophageal cancer (12).

The present study suggests that a minimal 5-microRNA classifier, consisting of hsa-miR-210, hsa-miR-182, hsa-miR-486-5p, hsa-miR-30a, and hsa-miR-140-3p, can distinguish malignant SCC tissues from normal tissues. This classifier is entirely new and has important molecular pathologic applications, because changes in the levels of the classifier species in tumor tissues may be detected as biomarkers. One particular powerful application of the 5-microRNA molecular classifier could be with TBLB, which has been explored as a method for early diagnosis of questionable lung infiltrations or densities (26). The diagnostic yield of TBLB is lower than that of open lung biopsy because the amount of lung tissue obtained is smaller. However, molecular detection methods require less amounts of tissue than traditional pathology and could improve the diagnostic accuracy of endoscopic procedures. The use of the microRNA classifier to improve the diagnostic yield after TBLB and avoid unnecessary surgery would be a significant advance. Although the TBLB tissue is scant, the RNA from the sample is sufficient for microRNA amplification and analysis, which could use U6 RNA from the same sample as an internal control for normalization. An alternative technical approach is to compare the expression of microRNAs from patients with common lung tissue control, which could be supplied in a kit form. It is not necessary to take the normal tissue from the same patient for analysis. A custom-based array chip could be designed to determine the 5-microRNA expression levels, using experiences learned from our previous array-based molecular diagnostic studies (27–29).

Furthermore, our study identified that the elevated expression of hsa-miR-31 correlates inversely with the survival of Chinese SCC patients. This single marker could be conveniently measured by qRT-PCR. It is notable that hsa-miR-31 is not a member of the 5-microRNA classifier identified in the same patient samples. The reason might be that the molecules involved in the initial occurrence of SCC and the later malignant effects are different. We also observed that TNM staging (I/II/III) was not consistently correlated with the SCC survival in our cohorts (Tables 2 and 3). Resection surgery cannot be successfully done on patients with TNM stage IV and part IIIb lung cancer; hence, samples of these stages were not available to us. The absence of these stage samples raises the possibility of unknown statistical bias in our Cox regression analysis. Yet, our findings suggest it may be beneficial to incorporate molecular markers into the lung cancer staging system to aid personalized treatment of patients, particularly at early to intermediate stages.

Recently, Yang and colleagues reported a profiling study on microRNA expression in SCC among 23 Chinese patients (30). Their conclusions were not concordant with our findings. Our study was the first comprehensive investigation of a statistically significant cohort of 148 Chinese patients with extensive clinicopathologic data. Interestingly, the 5 microRNAs comprising our SSC classifier have individually been reported to have roles in 3 other studies. Mascaux and colleagues reported a microRNA expression during bronchial squamous carcinogenesis (31). They used a total of 60 Caucasian biopsies, with 6 samples from each category: normal bronchial epithelium of nonsmokers, normal normofluorescent bronchial epithelial tissue of smokers, histologically normal but hypofluorescent bronchial epithelium of smokers, hyperplasia, metaplasia, mild, moderate, and severe dysplasia, carcinoma in situ, and SCC. The hsa-miR-486 was found dysregulated during lung carcinogenesis. Raponi and colleagues studied the prognosis of SCC using a cohort of 61 Caucasian SCC samples and 10 matched normal lung samples (32). hsa-miR-210 and hsa-miR-182 were among the microRNAs differentially expressed between SCC and normal lung samples in their cohort. They also reported a different microRNA from ours, hsa-miR-146b alone having prognostic value for SCC. In addition, hsa-miR-210 was one of the microRNA markers found in sputum by Xing and colleagues for early detection of SCC in Caucasians (33). The fact that some of our 5 microRNAs are common to 3 other studies with Caucasian patients is persuasive and it brings some confidence to the universality of our observations, but this assertion must be tested further.

MicroRNA-31 has been shown to be an oncogene in a murine cell line (ED-1 cell) through its repression of 2 tumor suppressors, PPP2R2A and LATS2 (34). Here, we examined the molecular role of hsa-miR-31 in a human SCC lung cancer cell line (SK-MES-1 cell) and found different results, which may be due to using cell lines from different species. Interestingly, our findings indicated that hsa-miR-31 suppressed the expression of DICER1 in the human SCC cell line, but no effect was found on PPP2R2A or LATS2. The expression of DICER1 has been shown to be downregulated in human lung cancer and its reduced expression correlates with shortened postoperative survival, which is consistent with our data (35). This novel finding that hsa-miR-31 can downregulate DICER1 is significant, considering DICER1 plays an important role in general microRNA expression and processing (3). Our results suggest that hsa-miR-31 may act as a negative regulator in the feedback loop controlling general microRNA expression and may have important implications for the systems biology of the cellular regulation of microRNA expression. The 3′-UTR homology to hsa-miR-31 was identical across all 3 genes (DICER1, PPP2R2A, and LATS2; Fig. 3A). It is particularly interesting why only DICER1 and not other genes (PPP2R2A and LATS2) were regulated by hsa-miR-31 in our study. One possible explanation is that microRNAs bind through partial sequence homology to the 3′-UTR of target genes, resulting at the same time in some degree of mRNA degradation and translation inhibition (4). Furthermore, the target gene sequences flanking the hsa-miR-31 binding site may affect the functions of hsa-miR-31. Grimson and colleagues also reported that the seed matches are not always sufficient for microRNA repression (36).

Significantly, the upregulation of hsa-miR-31 in human colorectal (37), liver (38), and head and neck tumors (39), as well as squamous cell carcinomas of the tongue (40), has been previously reported. Paradoxically, the downregulation of has-miR-31 is associated with malignant progression in human carcinomas of the breast (41), prostate (42), ovary (43), and stomach (44). These seemingly conflicting roles of hsa-miR-31 indicate that it may affect crucial events related to the progression of these tumors through complicated regulatory mechanisms (45). Further elucidation of the biological actions of hsa-miR-31 may prove significant to improve diagnosis and treatment of several types of human cancers. Our observation of its regulation of DICER1 activity may potentially explore wider biological functions of hsa-miR-31 in the development of other cancers.

In summary, we have carefully designed and investigated the microRNA expression profile in a large cohort of Chinese SCC patients, using cryopreserved archival tissues stored for at least 5 years for which extensive clinicopathologic data were available. We revealed a 5-microRNA classifier (hsa-miR-210, hsa-miR-182, hsa-miR-486-5p, hsa-miR-30a, and hsa-miR-140-3p) that can distinguish cancerous SCC lesions from adjacent normal tissues. We showed that a high level of hsa-miR-31 expression strongly correlated with a low survival for SCC patients. We showed that hsa-miR-31 directly targets the DICER1 3′-UTR and repressed the expression of DICER1 but not PPP2R2A or LATS2. The diagnostic 5-microRNA classifier and prognostic hsa-miR-31 should prove to be useful for the management of SCC among Chinese patients.

No potential conflicts of interest were disclosed.

The authors thank Meihua Xiong, Shujie Wang, and Xiaoyu Zhang for their excellent technical assistance.

This work was supported by grants from the National High Technology Research and Development Program of China (863 Program; No. 2006AA020707 and No. 2006AA02A401) and National Key Technology R&D Program (No. 2006BAI02A02).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Ferlay
J
,
Shin
HR
,
Bray
F
,
Forman
D
,
Mathers
C
,
Parkin
DM
. 
GLOBOCAN 2008, cancer incidence and mortality worldwide [Internet]
.
Lyon, France
:
International Agency for Research on Cancer
; 
2010
.
IARC CancerBase No. 10
.
Available from
: http://globocan.iarc.fr.
2.
Tanoue
LT
,
Detterbeck
FC
. 
New TNM classification for non-small-cell lung cancer
.
Expert Rev Anticancer Ther
2009
;
9
:
413
23
.
3.
Krol
J
,
Loedige
I
,
Filipowicz
W
. 
The widespread regulation of microRNA biogenesis, function and decay
.
Nat Rev Genet
2010
;
11
:
597
610
.
4.
Iorio
MV
,
Croce
CM
. 
MicroRNAs in cancer: small molecules with a huge impact
.
J Clin Oncol
2009
;
27
:
5848
56
.
5.
Croce
CM
. 
Causes and consequences of microRNA dysregulation in cancer
.
Nat Rev Genet
2009
;
10
:
704
14
.
6.
Yanaihara
N
,
Caplen
N
,
Bowman
E
,
Seike
M
,
Kumamoto
K
,
Yi
M
, et al
Unique microRNA molecular profiles in lung cancer diagnosis and prognosis
.
Cancer Cell
2006
;
9
:
189
98
.
7.
Yu
SL
,
Chen
HY
,
Chang
GC
,
Chen
HW
,
Singh
S
,
Cheng
CL
, et al
MicroRNA signature predicts survival and relapse in lung cancer
.
Cancer Cell
2008
;
13
:
48
57
.
8.
Markou
A
,
Tsaroucha
EG
,
Kaklamanis
L
,
Fotinou
M
,
Georgoulias
V
,
Lianidou
ES
. 
Prognostic value of mature microRNA-21 and microRNA-205 overexpression in non-small cell lung cancer by quantitative real-time RT-PCR
.
Clin Chem
2008
;
54
:
1696
704
.
9.
Voortman
J
,
Goto
A
,
Mendiboure
J
,
Sohn
JJ
,
Schetter
AJ
,
Saito
M
, et al
MicroRNA expression and clinical outcomes in patients treated with adjuvant chemotherapy after complete resection of non-small cell lung carcinoma
.
Cancer Res
2010
;
70
:
8288
98
.
10.
Patnaik
SK
,
Kannisto
E
,
Knudsen
S
,
Yendamuri
S
. 
Evaluation of microRNA expression profiles that may predict recurrence of localized stage I non-small cell lung cancer after surgical resection
.
Cancer Res
2010
;
70
:
36
45
.
11.
Hu
Z
,
Chen
X
,
Zhao
Y
,
Zhou
X
,
Gu
H
,
Xu
L
, et al
Serum microRNA signatures identified in a genome-wide serum microRNA expression profiling predict survival of non-small-cell lung cancer
.
J Clin Oncol
2010
;
28
:
1721
6
.
12.
Guo
Y
,
Chen
Z
,
Zhang
L
,
Zhou
F
,
Shi
S
,
Feng
X
, et al
Distinctive microRNA profiles relating to patient survival in esophageal squamous cell carcinoma
.
Cancer Res
2008
;
68
:
26
33
.
13.
Ding
J
,
Huang
S
,
Wu
S
,
Zhao
Y
,
Liang
L
,
Yan
M
, et al
Gain of miR-151 on chromosome 8q24.3 facilitates tumour cell migration and spreading through downregulating RhoGDIA
.
Nat Cell Biol
2010
;
12
:
390
9
.
14.
Watanabe
T
,
Takeda
A
,
Mise
K
,
Okuno
T
,
Suzuki
T
,
Minami
N
, et al
Stage-specific expression of microRNAs during Xenopus development
.
FEBS Lett
2005
;
579
:
318
24
.
15.
Thomson
JM
,
Parker
J
,
Perou
CM
,
Hammond
SM
. 
A custom microarray platform for analysis of microRNA gene expression
.
Nat Methods
2004
;
1
:
47
53
.
16.
Tusher
VG
,
Tibshirani
R
,
Chu
G
. 
Significance analysis of microarrays applied to the ionizing radiation response
.
Proc Natl Acad Sci U S A
2001
;
98
:
5116
21
.
17.
Efron
B
,
Gong
G
. 
A leisurely look at the bootstrap, the jackknife, and cross-validation
.
Am Stat
1983
;
37
:
36
48
.
18.
Griffiths-Jones
S
,
Grocock
RJ
,
van Dongen
S
,
Bateman
A
,
Enright
AJ
. 
miRBase: microRNA sequences, targets and gene nomenclature
.
Nucleic Acids Res
2006
;
34
:
D140
4
.
19.
John
B
,
Enright
AJ
,
Aravin
A
,
Tuschl
T
,
Sander
C
,
Marks
DS
. 
Human microRNA targets
.
PLoS Biol
2004
;
2
:
e363
.
20.
Lewis
BP
,
Shih
IH
,
Jones-Rhoades
MW
,
Bartel
DP
,
Burge
CB
. 
Prediction of mammalian microRNA targets
.
Cell
2003
;
115
:
787
98
.
21.
Krek
A
,
Grun
D
,
Poy
MN
,
Wolf
R
,
Rosenberg
L
,
Epstein
EJ
, et al
Combinatorial microRNA target predictions
.
Nat Genet
2005
;
37
:
495
500
.
22.
Chen
C
,
Ridzon
DA
,
Broomer
AJ
,
Zhou
Z
,
Lee
DH
,
Nguyen
JT
, et al
Real-time quantification of microRNAs by stem-loop RT-PCR
.
Nucleic Acids Res
2005
;
33
:
e179
.
23.
Tan
X
,
Zhou
F
,
Wan
J
,
Hang
J
,
Chen
Z
,
Li
B
, et al
Pin1 expression contributes to lung cancer: prognosis and carcinogenesis
.
Cancer Biol Ther
2010
;
9
:
111
9
.
24.
Patterson
TA
,
Lobenhofer
EK
,
Fulmer-Smentek
SB
,
Collins
PJ
,
Chu
TM
,
Bao
W
, et al
Performance comparison of one-color and two-color platforms within the MicroArray Quality Control (MAQC) project
.
Nat Biotechnol
2006
;
24
:
1140
50
.
25.
Shi
L
,
Campbell
G
,
Jones
WD
,
Campagne
F
,
Wen
Z
,
Walker
SJ
, et al
The MicroArray Quality Control (MAQC)-II study of common practices for the development and validation of microarray-based predictive models
.
Nat Biotechnol
2010
;
28
:
827
38
.
26.
Takayama
K
,
Nagata
N
,
Miyagawa
Y
,
Hirano
H
,
Shigematsu
N
. 
The usefulness of step sectioning of transbronchial lung biopsy specimen in diagnosing sarcoidosis
.
Chest
1992
;
102
:
1441
3
.
27.
Zhu
L
,
Jiang
G
,
Wang
S
,
Wang
C
,
Li
Q
,
Yu
H
, et al
Biochip system for rapid and accurate identification of mycobacterial species from isolates and sputum
.
J Clin Microbiol
2010
;
48
:
3654
60
.
28.
Guo
Y
,
Zhou
Y
,
Wang
C
,
Zhu
L
,
Wang
S
,
Li
Q
, et al
Rapid, accurate determination of multidrug resistance in M. tuberculosis isolates and sputum using a biochip system
.
Int J Tuberc Lung Dis
2009
;
13
:
914
20
.
29.
Li
CX
,
Pan
Q
,
Guo
YG
,
Li
Y
,
Gao
HF
,
Zhang
D
, et al
Construction of a multiplex allele-specific PCR-based universal array (ASPUA) and its application to hearing loss screening
.
Hum Mutat
2008
;
29
:
306
14
.
30.
Yang
Y
,
Li
X
,
Yang
Q
,
Wang
X
,
Zhou
Y
,
Jiang
T
, et al
The role of microRNA in human lung squamous cell carcinoma
.
Cancer Genet Cytogenet
2010
;
200
:
127
33
.
31.
Mascaux
C
,
Laes
JF
,
Anthoine
G
,
Haller
A
,
Ninane
V
,
Burny
A
, et al
Evolution of microRNA expression during human bronchial squamous carcinogenesis
.
Eur Respir J
2009
;
33
:
352
9
.
32.
Raponi
M
,
Dossey
L
,
Jatkoe
T
,
Wu
X
,
Chen
G
,
Fan
H
, et al
MicroRNA classifiers for predicting prognosis of squamous cell lung cancer
.
Cancer Res
2009
;
69
:
5776
83
.
33.
Xing
L
,
Todd
NW
,
Yu
L
,
Fang
H
,
Jiang
F
. 
Early detection of squamous cell lung cancer in sputum by a panel of microRNA markers
.
Mod Pathol
2010
;
23
:
1157
64
.
34.
Liu
X
,
Sempere
LF
,
Ouyang
H
,
Memoli
VA
,
Andrew
AS
,
Luo
Y
, et al
MicroRNA-31 functions as an oncogenic microRNA in mouse and human lung cancer cells by repressing specific tumor suppressors
.
J Clin Invest
2010
;
120
:
1298
309
.
35.
Karube
Y
,
Tanaka
H
,
Osada
H
,
Tomida
S
,
Tatematsu
Y
,
Yanagisawa
K
, et al
Reduced expression of Dicer associated with poor prognosis in lung cancer patients
.
Cancer Sci
2005
;
96
:
111
5
.
36.
Grimson
A
,
Farh
KK
,
Johnston
WK
,
Garrett-Engele
P
,
Lim
LP
,
Bartel
DP
. 
MicroRNA targeting specificity in mammals: determinants beyond seed pairing
.
Mol Cell
2007
;
27
:
91
105
.
37.
Motoyama
K
,
Inoue
H
,
Takatsuno
Y
,
Tanaka
F
,
Mimori
K
,
Uetake
H
, et al
Over- and under-expressed microRNAs in human colorectal cancer
.
Int J Oncol
2009
;
34
:
1069
75
.
38.
Wong
QW
,
Lung
RW
,
Law
PT
,
Lai
PB
,
Chan
KY
,
To
KF
, et al
MicroRNA-223 is commonly repressed in hepatocellular carcinoma and potentiates expression of Stathmin1
.
Gastroenterology
2008
;
135
:
257
69
.
39.
Liu
X
,
Chen
Z
,
Yu
J
,
Xia
J
,
Zhou
X
. 
MicroRNA profiling and head and neck cancer
.
Comp Funct Genomics
2009
:
837514
.
40.
Wong
TS
,
Liu
XB
,
Wong
BY
,
Ng
RW
,
Yuen
AP
,
Wei
WI
. 
Mature miR-184 as potential oncogenic microRNA of squamous cell carcinoma of tongue
.
Clin Cancer Res
2008
;
14
:
2588
92
.
41.
Valastyan
S
,
Reinhardt
F
,
Benaich
N
,
Calogrias
D
,
Szász
AM
,
Wang
ZC
, et al
A pleiotropically acting microRNA, miR-31, inhibits breast cancer metastasis
.
Cell
2009
;
137
:
1032
46
.
42.
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
:
1166
76
.
43.
Creighton
CJ
,
Fountain
MD
,
Yu
Z
,
Nagaraja
AK
,
Zhu
H
,
Khan
M
, et al
Molecular profiling uncovers a p53-associated role for microRNA-31 in inhibiting the proliferation of serous ovarian carcinomas and other cancers
.
Cancer Res
2010
;
70
:
1906
15
.
44.
Guo
J
,
Miao
Y
,
Xiao
B
,
Huan
R
,
Jiang
Z
,
Meng
D
, et al
Differential expression of microRNA species in human gastric cancer versus non-tumorous tissues
.
J Gastroenterol Hepatol
2009
;
24
:
652
57
.
45.
Valastyan
S
,
Weinberg
RA
. 
miR-31: a crucial overseer of tumor metastasis and other emerging roles
.
Cell Cycle
2010
;
9
:
2124
9
.