Surgical specimens of 18 ccRCC cases and adjacent normal renal tissues were collected between January 2013 and December 2014 in the Capital Medical University Affiliated Beijing Friendship Hospital (Beijing, China). The study was approved by the Research Ethics Boards of both Capital Medical University and Beijing Friendship Hospital. All subjects included in the protocol signed a declaration of informed consent. The tumor specimens were classified according to the 2010 American Joint Committee on Cancer (AJCC) staging system.

Total RNA was extracted from ccRCC and normal renal tissues using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. RNA quantity and purity were assessed using NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, MA, USA). The criteria for absorbance ratios are established at A260/A280 ≥1.8 and A260/A230 ≥1.5 indicating acceptable RNA purity. RNA integrity number (RIN) values were ascertained using Agilent RNA 6000 Nano assay (Agilent Technologies, Santa Clara, CA, USA). The criteria for RIN value is established at ≥5 indicating acceptable RNA integrity. Genomic DNA contamination was evaluated by gel electrophoresis. The RNA samples were stored at −80°C until analysis.

miRNA microarray analysis was performed by the Human miRNA OneArray^® v6 (Phalanx Biotech Group, Inc., Hsinchu, Taiwan). It contains triplicate 2539 unique miRNA probes from human (miRBase release v20) each printed in technical triplicate, and 114 experimental control probes. Small RNA was pre-enriched by Nanoseplook (Pall Corp., Port Washington, NY, USA) from 2.0 μg total RNA samples and labeled with miRNA ULS™ labeling kit (Kreatech Diagnostics, Amsterdam, The Netherlands). Labeled miRNA targets were hybridized to the human miRNA OneArray^® v6 with OneArray^® hybridization system. After 16 h of hybridization at 37°C, non-specific binding targets were washed away by three different washing steps (wash I 37°C for 5 min; wash II 37°C for 5 min, 25°C for 5 min; wash III rinse 20 times), and the slides were dried by centrifugation and scanned by an Axon 4000B scanner (Molecular Devices, Sunnyvale, CA, USA). The Cy5 fluorescent intensities of each probe were analyzed by GenePix 4.1 software (Molecular Devices). The raw intensity of each probe was processed by R program. Probes that passed the criteria were normalized by 75% median scaling normalization method. Normalized spot intensities were transformed to miRNA expression log₂ ratios between tumor tissues and adjacent normal tissues from ccRCC patients.

The TCGA data on miRNA, mRNA (RNASeq v2) and protein (RPPA) expression levels in ccRCC patients were obtained from https://www.synapse.org/. The miR-19a level, SMAD4 mRNA level, PTEN protein level and whole transcriptome data were used in the present study. Clinical data were downloaded from cBioPortal database (www.cbioportal.org).

The association between expression level of miR-19a and biological processes/microRNA target gene set was analyzed using Gene set enrichment analysis (GSEA v2.2, http://www.broad.mit.edu/gsea/). GSEA calculates a gene set enrichment score (ES) that estimates whether genes from pre-defined gene set (obtained from the Molecular Signatures Database, MSigDB, regulation of cell proliferation, http://software.broadinstitute.org; miR-19a target gene set, http://mirtarbase.mbc.nctu.edu.tw/) are enriched among the highest- (or lowest-) ranked genes or distributed randomly. Default settings were used. Thresholds for significance were determined by permutation analysis (1000 permutations). False discovery rate (FDR) was calculated. A gene set is considered significantly enriched when the FDR score is <0.05.

Human ccRCC cell line 769-P was purchased from the Cell Resource Center of Beijing Xiehe (Beijing, China) and cultivated in an incubator at 37°C with 5% CO₂. 769-P was maintained in Roswell Park Memorial Institute (RPMI)-1640 medium (Gibco, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS; HyClone Laboratories, Inc., Logan, UT, USA) as well as penicillin (100 U/ml; Thermo Fisher Scientific).

miR-19a mimics and its negative control (NC) sequences were designed and synthesized by Sangon Biotech, Co., Ltd. (Shanghai, China). Lipofectamine 2000 (Invitrogen) was used for miRNA transfection according to the manufacturer’s protocol. After 48 h of transfection, cells were used for the proliferation and western blot experiments.

Cells transfected with miR-19a mimics or NC sequence were plated at a density of 5×10³ cells/well onto 96-well plates at 37°C in an incubator with 5% CO₂. Cell proliferation was then assessed every 24 h using Cell Counting kit-8 (CCK-8; Sigma-Aldrich, St. Louis, MO, USA) according to standard protocol. For each sample at each time-point, 6 wells were analyzed. The experiment was repeated three times.

miRPath v3.0 was used to map the Gene Ontology (GO) of 39 differentially expressed miRNAs and unifying functional KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways of miR-19a in DIANA Tools database (http://snf-515788.vm.okeanos.grnet.gr/). To explore the functional annotation enrichment of the leading edge target genes of miR-19a in ccRCC, the GO and KEGG analyses were conducted using the WebGestalt (the Database for functional genomic, proteomic and large-scale genetic studies) online analysis tool (http://bioinfo.vanderbilt.edu/webgestalt/option.php) (15).

Search Tool for the Retrieval of Interacting Genes/Proteins (STRING; Search Tool for the Retrieval of Interacting Genes, http://string-db.org/) is a database of known and predicted protein interactions that may aid in the comprehensive description of cellular mechanisms and functions. The PPI network of the selected leading edge target genes of miR-19a in ccRCC was constructed using the STRING database.

The IHC-based protein expression data including high-resolution images were viewed and downloaded from the Human Protein Atlas web portal (www.proteinatlas.org).

Cells were collected and lysed in lysis buffer (Beijing CoWin Biotech, Co., Ltd., Beijing, China) in the presence of protease inhibitors for 30 min to extract total protein from cells transfected with miR-19a mimics or negative control, and protein levels were quantified using bicinchoninic acid assays (Beijing CoWin Biotech). Subsequently, 30 μg protein from each sample was loaded onto 10% sodium dodecyl sulfate (SDS) polyacrylamide gels and subjected to SDS-polyacrylamide gel electrophoresis (PAGE; Beijing CoWin Biotech). Protein was then transferred to nitrocellulose membranes (Sigma-Aldrich), which were blocked with bovine serum albumin blocking buffer (Invitrogen) for 1 h. Membranes were then incubated with primary antibodies targeting SMAD4 (1:2,000 dilution, cat. no. ab40759; Abcam), PTEN (1:1,000 dilution, cat. no. ab32199; Abcam), or GAPDH (1:5000 dilution, cat. no. ab70699; Abcam) overnight at 4°C, followed by incubation with goat anti-rabbit horseradish peroxidase (HRP)-conjugated secondary antibody (1:3,000 dilution, cat. no. CW0103; Beijing CoWin Biotech) for 1 h at room temperature. Detection was facilitated using an enhanced chemiluminescence kit and images were analyzed using ImageJ software (version 1.62; National Institute of Health, Bethesda, MD, USA).

Statistical analyses were performed using SPSS 19.0 (SPSS, Inc., Chicago, IL, USA) and Graphpad Prism 5 (Graphpad Software, Inc., San Diego, CA, USA). miRNA microarray data and TCGA data for paired samples were analyzed by paired sample t-test. TCGA and IHC data for unpaired samples were analyzed by independent samples t-test. The relationship between miR-19a expression level and clinical stages was analyzed by one-way ANOVA. The effect of miR-19a on the constituent ratio of clinical stages was explored by Pearson’s Chi-square test. The log-rank test for the generated Kaplan-Meier curve was conducted to evaluate the association between the expression level of miR-19a and the survival rate. Overall survival (OS) was defined as the time between the first surgery for primary ccRCC and death for any reason. Disease-free survival (DFS) was defined as the time between the first surgical resection and disease recurrence. Univariate and multivariate Cox proportional hazard regression analyses were used to estimate the prognostic significance of miR-19a in ccRCC patients. Growth curves were analyzed using a repeated-measures ANOVA followed by Bonferroni post hoc analysis. The correlation of miR-19a and target gene expression levels were analyzed by Pearson correlation analysis. P<0.05 was considered to indicate a statistically significant difference.

To identify the novel differentially expressed miRNAs related with ccRCC occurrence and development, miRNA microarray and TCGA_KIRC miRNA dataset were used in the present study. Eighteen pairs of ccRCC tissues and adjacent normal renal tissues were compared by miRNA microarray, 152 differentially expressed miRNAs (P<0.01; FC>3 and FC<0.33) were identified. Of which, 39 miRNAs are also consistently differentially expressed (P<0.05) in TCGA_KIRC miRNA seq v2 dataset. Cell proliferation is an important process of cancer development and progression (16) and proliferation related proteins predicted the poor outcome of cancer patients (17). To identify differentially expressed miRNAs correlated with cell proliferation, we searched GO annotation of 39 miRNAs. Among 39 dysregulated miRNAs, 23 miRNAs were annotated with cell proliferation. Notably, these 23 miRNAs were all upregulated in ccRCC tissues. miR-19a was one of rarely reported miRNAs in ccRCC and attracted our interest (Fig. 1A–C and Table I).

To further validate the abnormally upregulated expression level of miR-19a, we compared 68 pairs of ccRCC tissue specimens from TCGA database. Consistent with the results from unpaired samples, miR-19a was significantly upregulated in tumor tissue specimens compared with matched normal controls (Fig. 1D; P<0.0001).

To elucidate the role of miR-19a in ccRCC progression, the expression level of miR-19a in different stages of ccRCC patients was analyzed. The results revealed that with the increase of ccRCC stage, miR-19a expression level increased accordingly (P<0.0001; Fig. 2A), indicating the role of upregulated miR-19a in ccRCC progression and prognosis prediction. To further observe the role of miR-19a expression in ccRCC progression, constituent ratio analysis was performed in an expanded set of 236 primary ccRCC tissues and 71 normal samples. In the normal renal tissues, only 10/71 (14%) cases exhibited high expression level for miR-19a (>median). However, in I/II stage and III/IV stage of ccRCC tissues, 81/144 (56%) of cases and 62/92 (67%) of cases showed high expression level for miR-19a, respectively. The percentage with high level of miR-19a was drastically different between normal samples and ccRCC I/II or III/IV stage ccRCC tissues (P<0.01), suggesting that miR-19a was associated with ccRCC progression and might have prognostic significance for ccRCC patients (Fig. 2B). Thus, we further investigated the prognostic role of miR-19a expression level as a prognostic marker. The patients were divided into miR-19a high expression and miR-19a low expression groups. Kaplan-Meier curve showed that miR-19a high expression group had shorter OS and DFS than miR-19a low expression group in terms of survival duration (P<0.05; Fig. 2C and D), particularly in DFS. The result suggests that upregulated miR-19a may be a novel prognostic marker for ccRCC patients. Univariate and multivariate Cox regression analyses further revealed miR-19a could be an independent prognostic marker for ccRCC patients (Table II).

To specify which of the T, N and M stages was significantly related with miR-19a expression level, we next, respectively, investigated the relationship of miR-19a expression level with T, N and M stage. miR-19a expression level was most significantly related with T stage (P<0.05; Fig. 3A). T stage is closely related to tumor size and cell proliferation, revealing that miR-19a was significantly related with cell proliferation of ccRCC patients. GSEA was used to further verify the relationship of miR-19a expression level with cell proliferation in ccRCC patients. The ccRCC patients from TCGA dataset were divided into high and low miR-19a expression groups, and correlation between miR-19a expression level and cell proliferation gene set was analyzed. As shown in Fig. 3B, the gene set of cell proliferation was highly enriched in the miR-19a high expression group (FDR <0.05). The data suggest miR-19a expression level positively correlates with cell proliferation. Next, the ability of miR-19a in modulating ccRCC cell proliferation was analyzed in vitro. The results indicated that miR-19a overexpression significantly promoted the proliferation of 769-P cells by >50% at all time-points following 48 h of incubation, as compared with control cells (P<0.01). Moreover, no significant difference was observed between untransfected cells and negative control sequence-transfected cells (Fig. 3C). These data suggest that miR-19a promotes ccRCC cell proliferation.

miRNAs were reported to promote cancer occurence and progression through activating related signaling pathways (18). To better understand the underlying molecular mechanism by which miR-19a functions, we investigated miR-19a relevant KEGG pathways. Through the miRNA-targeted pathway union analysis, we found that miR-19a was involved in 49 significant KEGG pathways (P<0.05). These pathways mainly included multiple types of cancer related signaling pathways, especially the TGF-β signaling pathway, mTOR signaling pathway and PI3K-Akt signaling pathway (Fig. 4).

miRNAs exerted important functions in various biological processes via directly suppressing the expression of their target genes (19). However, for the same miRNA, target genes and involved pathways are different in different tissues (20). To find out the key and the specific target genes of miR-19a in ccRCC patients, we chose 28 target genes of miR-19a which as identified in the miRTarBase dataset and whose expression level value was listed in the TCGA KIRC dataset, investigated their correlation with miR-19a in ccRCC by GSEA. The results revealed that 19 target genes were significantly enriched in miR-19a low expression group, suggesting that they are the main target genes of miR-19a in ccRCC patients (Fig. 5A and Table III). These genes included SMAD4, PTEN, BCL2L11, SUZ12, PSAP, THBS1, KAT2B, VPS4B and BMPR2. Especially, SMAD4 and PTEN were the high ranked target genes for miR-19a. We further performed GO and KEGG pathway analysis for these target genes. We noted that the 19 genes were especially enriched in functions of metabolic process, biological regulation, macromolecular complex and protein binding (Fig. 5B). Another functional analysis of above target genes by KEGG annotation revealed that some signal transduction pathways, such as transforming growth factor (TGF)-β signaling pathway, pathway in cancer and p53 signaling pathway were significantly regulated by miR-19a (Fig. 5C). In addition, KEGG pathway analysis results also revealed that SMAD4 and PTEN were involved in many cell proliferation-related signaling pathways (data not shown). Since macromolecular complex and protein binding are the important molecular functions for target genes, STRING was used to analyze the interaction among the 19 target genes of miR-19a. SMAD4 and PTEN were found to play important roles in this network (Fig. 5D). In summary, data suggest that miR-19a promotes ccRCC occurrence, progression and poor prognosis mainly via suppressing SMAD4 and PTEN expression.

SMAD4 and PTEN are well-known tumor suppressors in multiple types of tumors (21,22). The abnormal downregulation of PTEN and SMAD4 protein plays an important role in the malignant transformation of tumors (23,24). In ccRCC tissues (n=11), SMAD4 and PTEN protein expression levels were significantly downregulated (Fig. 6A, P=0.0023; P<0.0001, respectively) compared with normal renal tissues from the Human Protein Atlas Database (n=3). To further verify the expression levels, SMAD4 and PTEN were regulated by miR-19a from clinical aspect, the correlation analyses between expression levels of SMAD4/PTEN and miR-19a were done in ccRCC tissues. The results showed that the expression level of miR-19a was negatively correlated with the expression level of both SMAD4 and PTEN (Fig. 6B, R²=0.015 and 0.11, respectively). To validate that miR-19a can directly suppress the expression of SMAD4 and PTEN, miR-19a mimics are transfected into 769-P cells and the expression levels of SMAD4 and PTEN were detected. The results demonstrated that the expression levels of both SMAD4 and PTEN were reduced with miR-19a overexpression in 769-P cells (Fig. 6C). These data suggest that miR-19a suppresses the expression of SMAD4 and PTEN in ccRCC.