Abstract
Analyses of our previously determined microRNA (miRNA) expression signature of renal cell carcinoma (RCC) and The Cancer Genome Atlas (TCGA) database revealed that both strands of the pre-miR-532-duplex-miR-532-5p (the guide strand) and miR-532-3p (the passenger strand)- are closely associated with poor prognosis of RCC patients (P = 0.0411 and P = 0.022, respectively). In this study we investigated the functional significance of these miRNAs and identified gene targets involved in RCC pathogenesis. Ectopic expression of these miRNAs significantly attenuated the malignant phenotypes including proliferation, migration and invasion of two RCC cell lines, 786-O and A498. A combination of genome-wide gene expression and in silico database analyses revealed 36 and 34 genes as putative target oncogenes regulated by miR-532-5p and miR-532-3p, respectively, in RCC cells. Among these targets, expression of aquaporin9 (AQP9), a water channel protein, was directly regulated by both miR-532-5p and miR-532-3p, and high expression levels of AQP9 were significantly associated with poor prognosis of RCC patients (P = 2.03e-05). Multivariate analysis indicated that AQP9 expression is an independent prognostic factor for RCC patients. Aberrant AQP9 expression at both the gene and protein level was detected in RCC clinical specimens. siRNA-mediated knockdown of AQP9 by si-AQP9 inhibited the malignant phenotypes of RCC cells. Rescue assays of AQP9 overexpression showed that the miR-532/AQP9 axis was closely involved in RCC oncogenesis. The identification of antitumor miRNAs and their targets will contribute to an increased understanding of the molecular pathogenesis of RCC.
Keywords: microRNA, antitumor, miR-532-5p, miR-532-3p, passenger strand, renal cell carcinoma, AQP9
Introduction
Renal cell carcinoma (RCC) accounts for 2-3% of all adult cancers and is the seventh and ninth most common cancer for men and women, respectively [1]. An estimated 209,000 new cases and 102,000 deaths due to all RCCs occur annually worldwide [1]. There are many RCC types including clear cell RCC, papillary RCC, chromophobe RCC and collecting duct RCC. Each type of RCC has distinct characteristics and is derived from a different cell lineage. The most common type of RCC is clear cell RCC (ccRCC), that represents approximately 75%-85% of RCC [2].
The 5-year overall survival rate of patients with early-stage RCC is approximately 60%, whereas the rate for patients with metastasis or advanced-stage is far lower, at less than 10% [1]. Between 30% and 40% of patients with early-stage RCC develop local recurrence or distant metastasis [2]. Several recently developed treatment strategies for RCC, such as mTOR-targeted agents, VEGF-tyrosine kinase inhibitors and immune checkpoint inhibitors, have slightly improved overall survival rates for patients with metastatic disease [3]. Despite these improvements, the 5-year survival of advanced RCC continues to be around 20% [3]. Elucidation of metastatic pathways and therapeutic targets will be essential for achieving improved outcomes for RCC patients.
Advanced cancer-genome technologies have highlighted the important role that noncoding RNAs play in disease pathogenesis. Noncoding, single-stranded microRNAs (miRNAs) have 19 to 22 nucleotides and can fine-tune expression of both protein coding and non-coding genes by repressing translation or cleaving RNA transcripts in a sequence-dependent manner [4]. A single miRNA can regulate a large number of protein-coding and noncoding RNA transcripts in relevant cells. As such, aberrant expression of miRNAs can have broad-ranging effects on RNA networks within cells. Numerous studies have revealed that dysregulated miRNA expression can disrupt tightly controlled RNA networks. Such disruptions are closely related to cancer cell development, metastasis and drug resistance [4].
Based on miRNA expression signatures in RCC, we sequentially identified antitumor miRNAs and the oncogenic targets they directly control, including miR-101 (target UHRF1), miR-10a-5p (SKA1), miR-26a/miR-26b (LOXL2 and PLOD2) and miR-451a (PMM2) [5-8]. We recently showed that both strands of the pre-miRNAs pre-miR-144 and pre-miR-455 acted as antitumor miRNAs in RCC cells and the oncogenic genes they target are closely involved in RCC pathogenesis [9,10]. The conventional theory for the biological function of miRNA suggested that the guide strands of miRNA can control expression of target genes, whereas passenger strands are degraded and have no function [11]. However, our studies revealed that several miRNA passenger strands can indeed regulate target gene expression and the aberrant expression of these miRNAs is involved in RCC oncogenesis.
Analyses of our original miRNA expression signature for RCC and The Cancer Genome Atlas (TCGA) database revealed that both miR-532-5p (the guide strand) and miR-532-3p (the passenger strand) are closely associated with poor prognosis of RCC patients (P = 0.0411 and P = 0.022, respectively). Here we investigated the functional significance of these miRNAs in terms of the oncogenes they target and their role in RCC pathogenesis.
Materials and methods
Clinical RCC specimens and RCC cell lines
A total of 23 clinical RCC tissue samples were obtained from patients that underwent total nephrectomy at Chiba University Hospital between 2008 and 2015 (Table 1). No patient had metastatic sites at the time of surgery. All patients in this study signed informed consent and the present study protocol was approved by the Institutional Review Board of Chiba University (acceptance number: 484). We used the RCC cell lines 786-0 and A498 that were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA).
Table 1.
Characteristics of 23 patients with non-metastatic clear cell RCC
| No. | Age at operation | Sex | Tumor Grade | pT | Remarks |
|---|---|---|---|---|---|
| 1 | 71 | F | G2 | T1a | qRT-PCR |
| 2 | 74 | M | G1>G2 | T1a | qRT-PCR |
| 3 | 59 | M | G3>G2 | T1b | qRT-PCR |
| 4 | 52 | M | G2>G3>G1 | T1a | qRT-PCR |
| 5 | 64 | M | G2>G3 | T1b | qRT-PCR |
| 6 | 67 | M | G2>G3>G1 | T3a | qRT-PCR |
| 7 | 59 | M | G3 | T3a | qRT-PCR |
| 8 | 61 | M | G2>G1 | T1a | qRT-PCR |
| 9 | 73 | M | G1>>G3 | T2a | qRT-PCR |
| 10 | 77 | M | G1>G2 | T1b | qRT-PCR |
| 11 | 77 | M | G2>G1 | T3a | qRT-PCR |
| 12 | 66 | M | G2>G1 | T1a | qRT-PCR |
| 13 | 51 | F | G2>G1>G3 | T3a | qRT-PCR |
| 14 | 47 | M | G2>G1 | T1b | qRT-PCR |
| 15 | 78 | M | G2>G1>>G3 | T1b | qRT-PCR |
| 16 | 44 | M | G2>G1 | T1a | qRT-PCR |
| 17 | 57 | M | G1>G2 | T1a | qRT-PCR |
| 18 | 54 | M | G2>G1 | T3a | qRT-PCR |
| 19 | 72 | F | G1>>G2 | T1b | qRT-PCR |
| 20 | 70 | M | G3>G2 | T3a | qRT-PCR |
| 21 | 71 | M | G2>G1 | T1a | IHC |
| 22 | 66 | F | G2 | T1b | IHC |
| 23 | 68 | M | G1 | T1b | IHC |
RCC, renal cell carcinoma; F, female; M, male; qRT-PCR, quantitative reverse transcription polymerase chain reaction; IHC, immunohistochemistry.
Transfection of mature miRNA, siRNA or plasmid vectors
We used the following RNAs in this study: pre-miR miRNA precursors (hsa-miR-532-5p, assay ID: PM11553; hsa-miR-532-3p, assay ID: PM12824; Applied Biosystems, Foster City, CA, USA), negative control miRNA (assay ID: AM 17111; Applied Biosystems) and siRNA (Stealth Select RNAi siRNA; si-AQP9, P/N: HSS100615 and HSS100617; Invitrogen, Carlsbad, CA, USA). The AQP9 plasmid vector was provided by ORIGENE (cat. no. SC113060; Rockville, MD, USA). Transfections were carried out using previously described procedures [8]. miRNAs and siRNAs were incubated with Opti-MEM (Invitrogen) and Lipofectamine RNAi Max transfection reagent at 10 nM (Invitrogen). Plasmid vectors were incubated with Opti-MEM and Lipofectamine 3000 reagent (Invitrogen) for forward transfection.
Quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR)
TaqMan probes and AQP9 primers (P/N: Hs01033361_m1; Applied Biosystems) were assay-on-demand gene expression products. qRT-PCR for miR-532-5p (P/N: 001518; Applied Biosystems) and miR-532-3p (P/N: 002355; Applied Biosystems) was used to validate miRNA expression. To normalize the data for analysis of mRNAs and miRNAs, GUSB (P/N: Hs99999908_m1; Applied Biosystems) and RNU48 (assay ID: 001006; Applied Biosystems) were used. PCR quantification was carried out as described previously [12].
Cell proliferation, migration, and invasion assays
Cell proliferation activity was determined using the XTT assay with the Cell Proliferation Kit II (Sigma-Aldrich, St. Louis, MO, USA). Cell migration was assessed using wound healing assays. Cell invasion activity was determined using modified Boyden chambers containing Matrigel-coated Transwell membrane filter inserts.
Western blotting
Western blotting was performed with polyclonal anti-AQP9 antibodies (1:200 dilution; SAB4301752; Sigma-Aldrich). We used anti-glyceraldehyde 3-phosphate dehydrogenase (GAPDH) antibodies (1:10,000 dilution; ab8245; Abcam, Cambridge, UK) as a control.
Incorporation of miR-532-5p and miR-532-3p into the RISC by Ago2 immunoprecipitation
A498 cells were transfected with 10 nM miRNA by reverse transfection. After 72 h, immunoprecipitation was performed using an Ago2 miRNA isolation kit (Wako, Osaka, Japan). Expression levels of miR-532-5p and miR-532-3p were analyzed by qRT-PCR. miRNA data were normalized to miR-26a expression (P/N: 000405; Applied Biosystems), which was not affected by miR-532-5p and miR-532-3p.
Identification of candidate target genes regulated by miR-532-5p and miR-532-3p
Candidate target genes regulated by miR-532-5p and miR-532-3p were identified using a combination of in silico and genome-wide gene expression analyses and listed in the TargetScan database (release 7.0) in a sequence-dependent manner (http://www.targetscan.org/vert_70/). Upregulated genes in RCC were identified from public data in the Gene Expression Omnibus (GEO; accession number: GSE36895) and we narrowed down the candidate genes. Gene expression was analyzed with our own oligo microarray data analyses (Human GE 60K; Agilent Technologies) that were deposited into the GEO (on June 14th, 2018; http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE115800.
Dual-luciferase reporter assay
The wild-type sequence of the AQP9 3’-UTR was inserted between the SgfI and PmeI restriction sites in the 3’-UTR of the hRluc gene in the psiCHECK-2 vector (C8021; Promega, Madison, WI, USA). We used sequences that were deleted in the miR-532-5p (position 1604-1610) and miR-532-3p (position 935-941) target sites. psiCHECK-2 vector was used as a cloning vector for the synthesized DNA [8].
Immunohistochemistry
Immunohistochemistry procedures were performed according to a previously described method [12]. Clinical tissue sections were incubated overnight at 4°C with anti-AQP9 antibodies diluted 1:20 (SAB4301752; Sigma-Aldrich).
Analysis of genes downstream of AQP9
To investigate AQP9-regulated pathways in RCC cells we assessed gene expression changes in A498 cells transfected with si-AQP9. Microarray data were used for expression profiling of si-AQP9 transfectants. The data were deposited into the GEO on December 4, 2018 (accession number: GSE123317).
Analysis of clinical significance of the miR-532-duplex and AQP9
We examined the clinical importance of miRNAs and genes in RCC patients using the RNA sequencing database in TCGA (https://tcga-data.nci.nih.gov/tcga/). The gene expression and clinical information were obtained from cBioportal (http://www.cbioportal.org/ and the provisional data were downloaded on September 1, 2018 [13-15].
Statistical analysis
Statistical analyses involving two or three variables and numerical values were analyzed by Bonferroni-adjusted Mann-Whitney U tests. Spearman’s rank tests were used to analyze correlations between expression levels. Expert StatView software (version 5.0, SAS Institute Inc., Cary, NC, USA) was used for these analyses. Multivariate analysis with JMP Pro 13 was used to analyze prognostic factors for patient survival.
Results
Analysis of miR-532-5p and miR-532-3p expression levels in clinical tissue samples
In the human genome, miR-532 is located on chromosome Xp11.23. The mature sequences of miR-532-5p and miR-532-3p are 5’-caugccuugaguguaggaccgu-3’ and 5’-ccucccacacccaaggcuugca-3’, respectively. The expression levels of miR-532-5p and miR-532-3p were significantly decreased in cancer tissues compared with those in adjacent noncancerous tissues (P < 0.0001; Figure 1A, 1B). In addition, Spearman’s rank test showed a strong positive correlation between miR-532-5p and miR-532-3p expression levels in RCC specimens (R = 0.776, P < 0.0001; Figure 1C).
Figure 1.
miR-532-duplex expression, relation to prognosis, and function. A-C. Expression levels of miR-532-5p and miR-532-3p in RCC clinical specimens. RNU48 was used as an internal control. Expression levels of miR-532-5p and miR-532-3p were positively correlated by Spearman’s rank test. D, E. Based on TCGA database, low expression levels of both miR-532-5p and miR-532-3p were significantly associated with poor prognosis. F-H. Cell proliferation, migration and invasion activities. *P < 0.01, **P < 0.0001.
Clinical significance and functional analysis of miR-532-5p and miR-532-3p
From TCGA database, low expression levels of miR-532-5p and miR-532-3p were significantly associated with a shorter survival period for RCC patients (P = 0.0411 and P = 0.022, respectively, Figure 1D, 1E). In both 786-O and A498 cells, restoration of miR-532-5p and miR-532-3p expression indicated that both miRNAs could suppress cancer cell proliferation, migration and invasion activities (Figure 1F-H).
Incorporation of miR-532-5p and miR-532-3p into the RISC
We next performed immunoprecipitation with antibodies targeting Ago2, which plays a pivotal role in the uptake of miRNAs into the RNA-induced silencing complex (RISC) (Figure 2A). After transfection of A498 cells with miR-532-5p and immunoprecipitation by anti-Ago2 antibodies, miR-532-5p levels in the immunoprecipitates were significantly higher than those of mock- or miR-control-transfected cells and those of miR-532-3p-transefected cells (P < 0.0001; Figure 2B). Similarly, after miR-532-3p transfection, substantial amounts of miR-532-3p were detected in Ago2 immunoprecipitates (P < 0.0001; Figure 2C).
Figure 2.
Incorporation of miR-532-5p and miR-532-3p into the RISC. A. Schematic diagram illustrating Ago2 immunoprecipitation. B, C. Amount of miR-532-5p and miR-532-3p in Ago2 immunoprecipitates of cell lysates from cells transfected transfection with miR-532-5p or miR-532-3p. The amount shown is relative to mock-transfected cells. *P < 0.0001.
Identification of candidate target genes of miR-532-5p and miR-532-3p regulation
We next searched for genes that had putative target sites for miR-532-5p and miR-532-3p in their 3’-UTR and that showed upregulated expression levels (Fold Change > 2) in RCC tissues and downregulated expression in RCC cells transfected with miR-532-5p or miR-532-3p (Log2 ratio < -1) (Figure 3). Using this search strategy, we identified 36 and 34 genes as candidate target genes of miR-532-5p and miR-532-3p, respectively (Tables 2, 3). Among these candidate genes, RCC patients with high expression of 10 genes and 11 genes had significantly poor prognosis from TCGA database (Figures 4, 5). We focused on aquaporin9 (AQP9), which is targeted by both miR-532-5p and miR-532-3p, because aquaporins affect not only water and small molecule permeability but are also implicated in the development of several types of cancers.
Figure 3.
Search strategy to identify miR-532-5p and miR-532-3p target genes. The venn diagram represents genes identified through a search of the TargetScan database to identify genes carrying miR-532 binding sites as well as genes that have up- or down-regulated expression in the GEO database for RCC patients and in miR-532-transfected cells.
Table 2.
Candidate miR-532-5p target genes in RCC
| Gene Symbol | Gene Name | Entrez Gene ID | Cytoband | GEO expression data Fold-Change (Tumor/Normal) | miR-532-5p transfection in 786-O (Log2 ratio) | Binding Sites Count | TCGA analysis for OS (high vs low expression: p value) |
|---|---|---|---|---|---|---|---|
| FAM64A | Family with sequence similarity 64N, member A | 54478 | hs|17p13.2 | 2.400 | -1.450 | 1 | 1.79E-07 |
| CEP55 | Centrosomal protein 55 kDa | 55165 | hs|10q23.33 | 4.202 | -1.891 | 1 | 6.94E-07 |
| AQP9 | Aquaporin 9 | 366 | hs|15q21.3 | 2.077 | -1.470 | 1 | 2.03E-05 |
| DEPDC1 | DEP domain containing 1 | 55635 | hs|1p31.2 | 2.607 | -2.934 | 1 | 0.000111 |
| TCF19 | Transcription factor 19 | 6941 | hs|6p21.33 | 3.277 | -1.199 | 1 | 0.000554 |
| MKI67 | Antigen identified by monoclonal antibody Ki-67 | 4288 | hs|10q26.2 | 2.039 | -1.047 | 1 | 0.00106 |
| PAQR4 | Progestin and adipoQ receptor family member IV | 124222 | hs|16p13.3 | 5.134 | -1.173 | 1 | 0.00152 |
| CENPK | Centromere protein K | 64105 | hs|5q12.3 | 4.442 | -1.556 | 1 | 0.00234 |
| KIAA0101 | KIAA0101 | 9768 | hs|15q22.31 | 3.359 | -1.237 | 2 | 0.00258 |
| CSF1 | Colony stimulating factor 1 (macrophage) | 1435 | hs|1p13.3 | 2.021 | -1.547 | 0 | 0.027 |
| KIAA1715 | KIAA1715 | 80856 | hs|2q31.1 | 2.170 | -2.271 | 2 | 0.0592 |
| BCAT1 | Branched chain amino-acid transaminase 1, cytosolic | 586 | hs|12p12.1 | 3.076 | -2.187 | 3 | 0.1 |
| MEGF6 | Multiple EGF-like-domains 6 | 1953 | hs|1p36.32 | 2.113 | -2.224 | 1 | 0.151 |
| MCTP2 | Multiple C2 domains, transmembrane 2 | 55784 | hs|15q26.2 | 2.072 | -1.484 | 2 | 0.191 |
| NOTCH4 | Notch 4 | 4855 | hs|6p21.32 | 2.169 | -1.025 | 1 | 0.21 |
| RASSF5 | Ras association (RalGDS/AF-6) domain family member 5 | 83593 | hs|1q32.1 | 2.840 | -2.636 | 2 | 0.273 |
| SIPA1L2 | Signal-induced proliferation-associated 1 like 2 | 57568 | hs|1q42.2 | 2.462 | -1.356 | 1 | 0.28 |
| CHSY3 | Chondroitin sulfate synthase 3 | 337876 | hs|5q23.3 | 4.220 | -1.391 | 1 | 0.318 |
| ABCA12 | ATP-binding cassette, sub-family A (ABC1), member 12 | 26154 | hs|2q35 | 3.975 | -3.277 | 1 | 0.32 |
| PLK2 | Polo-like kinase 2 | 10769 | hs|5q11.2 | 5.045 | -1.428 | 1 | 0.327 |
| HHIPL1 | HHIP-like 1 | 84439 | hs|14q32.2 | 2.044 | -1.318 | 1 | 0.343 |
| LZTS1 | Leucine zipper, putative tumor suppressor 1 | 11178 | hs|8p21.3 | 2.138 | -2.116 | 1 | 0.426 |
| SLAMF7 | SLAM family member 7 | 57823 | hs|1q23.3 | 4.181 | -2.447 | 2 | 0.44 |
| ERO1L | ERO1-like (S. cerevisiae) | 30001 | hs|14q22.1 | 2.829 | -2.441 | 0 | 0.444 |
| NCF2 | Neutrophil cytosolic factor 2 | 4688 | hs|1q25.3 | 3.432 | -2.254 | 1 | 0.495 |
| GBP1 | Guanylate binding protein 1, interferon-inducible | 2633 | hs|1p22.2 | 3.410 | -3.062 | 1 | 0.523 |
| ITGA4 | Integrin, alpha 4 (antigen CD49D, alpha 4 subunit of VLA-4 receptor) | 3676 | hs|2q31.3 | 2.336 | -2.043 | 1 | 0.573 |
| DOPEY2 | Dopey family member 2 | 9980 | hs|21q22.12 | 2.118 | -1.689 | 1 | 0.638 |
| MCTP1 | Multiple C2 domains, transmembrane 1 | 79772 | hs|5q15 | 3.092 | -1.179 | 0 | 0.749 |
| CD84 | CD84 molecule | 8832 | hs|1q23.3 | 3.645 | -1.970 | 1 | 0.781 |
| DCLK1 | Doublecortin-like kinase 1 | 9201 | hs|13q13.3 | 3.633 | -1.286 | 2 | 0.804 |
| CMPK2 | Cytidine monophosphate (UMP-CMP) kinase 2, mitochondrial | 129607 | hs|2p25.2 | 2.748 | -1.538 | 1 | 0.829 |
| BHLHE41 | Basic helix-loop-helix family, member e41 | 79365 | hs|12p12.1 | 9.461 | -1.129 | 2 | 0.895 |
| ARHGAP42 | Rho GTPase activating protein 42 | 143872 | hs|11q22.1 | 2.075 | -1.514 | 2 | 0.0000485* |
| SERTAD2 | SERTA domain containing 2 | 9792 | hs|2p14 | 2.183 | -1.375 | 1 | 0.00549* |
| FAR2 | Fatty acyl CoA reductase 2 | 55711 | hs|12p11.22 | 2.113 | -1.022 | 1 | 0.0136* |
Poor prgonosis in patients with low gene expression.
Table 3.
Candidate miR-532-3p target genes in RCC
| Gene Symbol | Gene Name | Entrez Gene ID | Cytoband | GEO expression data Fold-Change (Tumor/Normal) | miR-532-3p transfection in 786-O Binding Sites Count (Log2 ratio) | TCGA analysis for OS (high vs low expression: p value) | |
|---|---|---|---|---|---|---|---|
| C1orf216 | Chromosome 1 open reading frame 216 | 127703 | hs|1p34.3 | 2.220 | -1.493 | 1 | 5.69E-06 |
| RRM2 | Ribonucleotide reductase M2 | 6241 | hs|2p25.1 | 4.814 | -1.939 | 4 | 1.91E-05 |
| AQP9 | Aquaporin 9 | 366 | hs|15q21.3 | 2.077 | -1.455 | 1 | 2.03E-05 |
| PLXDC1 | Plexin domain containing 1 | 57125 | hs|17q12 | 3.144 | -1.534 | 2 | 0.00186 |
| SLC6A1 | Solute carrier family 6 (neurotransmitter transporter), member 1 | 6529 | hs|3p25.3 | 2.527 | -1.548 | 1 | 0.00285 |
| CHST2 | Carbohydrate (N-acetylglucosamine-6-O) sulfotransferase 2 | 9435 | hs|3q24 | 2.232 | -1.381 | 1 | 0.00288 |
| PTP4A3 | Protein tyrosine phosphatase type IVA, member 3 | 11156 | hs|8q24.3 | 2.360 | -1.542 | 1 | 0.00694 |
| SERPING1 | Serpin peptidase inhibitor, clade G (C1 inhibitor), member 1 | 710 | hs|11q12.1 | 2.015 | -1.244 | 1 | 0.0169 |
| SLC7A11 | Solute carrier family 7 (anionic amino acid transporter light chain, xc- system), member 11 | 23657 | hs|4q28.3 | 2.678 | -1.125 | 1 | 0.0234 |
| CSF1 | Colony stimulating factor 1 (macrophage) | 1435 | hs|1p13.3 | 2.021 | -1.323 | 3 | 0.027 |
| HCG27 | HLA complex group 27 (non-protein coding) | 253018 | hs|6p21.33 | 2.421 | -1.053 | 2 | 0.0447 |
| CARD11 | Caspase recruitment domain family, member 11 | 84433 | hs|7p22.2 | 2.190 | -1.698 | 1 | 0.0556 |
| KIAA1462 | KIAA1462 | 57608 | hs|10p11.23 | 2.183 | -1.551 | 1 | 0.0684 |
| SCD | Stearoyl-CoA desaturase (delta-9-desaturase) | 6319 | hs|10q24.31 | 6.464 | -2.610 | 1 | 0.129 |
| ADAMTS2 | ADAM metallopeptidase with thrombospondin type 1 motif, 2 | 9509 | hs|5q35.3 | 2.863 | -1.150 | 2 | 0.132 |
| DOCK2 | Dedicator of cytokinesis 2 | 1794 | hs|5q35.1 | 5.262 | -1.429 | 2 | 0.172 |
| PGBD5 | PiggyBac transposable element derived 5 | 79605 | hs|1q42.13 | 3.517 | -1.212 | 1 | 0.176 |
| RUNX3 | Runt-related transcription factor 3 | 864 | hs|1p36.11 | 3.644 | -1.263 | 2 | 0.213 |
| MET | Met proto-oncogene | 4233 | hs|7q31.2 | 2.553 | -1.684 | 1 | 0.224 |
| TTYH3 | Tweety family member 3 | 80727 | hs|7p22.3 | 2.138 | -1.423 | 1 | 0.247 |
| DTX3L | Deltex 3-like (Drosophila) | 151636 | hs|3q21.1 | 2.253 | -1.338 | 2 | 0.272 |
| PSMB9 | Proteasome (prosome, macropain) subunit, beta type, 9 | 5698 | hs|6p21.32 | 3.722 | -1.137 | 4 | 0.367 |
| NCAPG2 | Non-SMC condensin II complex, subunit G2 | 54892 | hs|7q36.3 | 2.127 | -1.127 | 1 | 0.385 |
| MEF2C | Myocyte enhancer factor 2C | 4208 | hs|5q14.3 | 2.693 | -1.276 | 1 | 0.472 |
| PFKFB3 | 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 | 5209 | hs|10p15.1 | 2.217 | -1.285 | 1 | 0.507 |
| RTP4 | Receptor (chemosensory) transporter protein 4 | 64108 | hs|3q27.3 | 3.204 | -1.335 | 1 | 0.545 |
| TRIM9 | Tripartite motif containing 9 | 114088 | hs|14q22.1 | 3.764 | -2.307 | 3 | 0.552 |
| TMEM92 | Transmembrane protein 92 | 162461 | hs|17q21.33 | 2.788 | -2.151 | 4 | 0.635 |
| RNF145 | Ring finger protein 145 | 153830 | hs|5q33.3 | 2.430 | -1.203 | 1 | 0.751 |
| RASSF2 | Ras association (RalGDS/AF-6) domain family member 2 | 9770 | hs|20p13 | 6.147 | -2.224 | 3 | 0.911 |
| BICD1 | Bicaudal D homolog 1 (Drosophila) | 636 | hs|12p11.21 | 2.423 | -1.079 | 1 | 0.955 |
| KDR | Kinase insert domain receptor (a type III receptor tyrosine kinase) | 3791 | hs|4q12 | 2.040 | -1.301 | 1 | 0.00000477* |
| ZBTB42 | Zinc finger and BTB domain containing 42 | 100128927 | hs|14q32.33 | 2.091 | -2.218 | 3 | 0.000469* |
| DIRAS2 | DIRAS family, GTP-binding RAS-like 2 | 54769 | hs|9q22.2 | 6.202 | -1.019 | 4 | 0.00119* |
Poor prgonosis in patiets with low gene expression.
Figure 4.
Kaplan-Meier analysis of candidate targets of miR-532-5p from TCGA database. Kaplan-Meier plots to compare overall survival rates of patients that had high (red lines) and low (blue lines) expression of 10 genes regulated by miR-532-5p (patients were divided in half).
Figure 5.
Kaplan-Meier analysis of candidate targets of miR-532-3p from TCGA database. Kaplan-Meier plots to compare overall survival rates of patients that had high (red lines) and low (blue lines) expression of 11 genes regulated by miR-532-5p (patients were divided in half).
Direct regulation of AQP9 expression by miR-532-5p and miR-532-3p
Levels of AQP9 mRNA and protein were significantly suppressed following transfection of 786-0 and A498 cells with miR-532-5p or miR-532-3p compared with mock-transfected cells or those transfected with miR-control (Figure 6A, 6B).
Figure 6.
Direct regulation of AQP9 expression by miR-532-5p and miR-532-3p. A. AQP9 mRNA expression levels 48 h after transfection of 786-0 or A498 RCC cells with 10 nM miR-532-5p or miR-532-3p. GAPDH was used as an internal control. B. Protein expression of AQP9 72 h after transfection with miR-532-5p or miR-532-3p. GAPDH was used as a loading control. C. miR-532-5p and miR-532-3p binding sites in the AQP9 mRNA 3’-UTR. D. Dual luciferase reporter assays with vectors encoding putative miR-532-5p and miR-532-3p target sites in the wild-type AQP9 3’-UTR and a 3’-UTR with the target sites deleted (Deletion). Normalized data were calculated as the ratio of Renilla/firefly luciferase activities. *P < 0.0001, **P < 0.01.
The TargetScan Human database indicated the presence of binding sites for both miR-532-5p (position 1604-1610) and miR-532-3p (position 935-941) in the AQP9 3’-UTR. Thus, we performed luciferase reporter assays with a vector including these sequences to assess whether miR-532-5p and miR-532-3p directly regulated AQP9 expression in a sequence-dependent manner. Cotransfection with miR-532-5p or miR-532-3p with vectors significantly suppressed luciferase activity in comparison with those in mock and miR-control transfectants (Figure 6C, 6D).
Expression of AQP9 in clinical specimens
AQP9 mRNA expression levels were significantly upregulated in cancer tissues compared with that in adjacent noncancerous tissues (P = 0.0069) (Figure 7A). Spearman’s rank test revealed a negative correlation between AQP9 expression and that of the miR-532-duplex (5p; P = 0.0077, R = -0.427 and 3p; P = 0.0027, R = -0.48, respectively; Figure 7B, 7C). Immunostaining for AQP9 in RCC clinical specimens indicated that AQP9 was strongly overexpressed in cancer lesions compared with adjacent noncancerous tissues at the same staining intensity (Figure 7D).
Figure 7.
Expression of AQP9 in clinical specimens. A. Expression levels of mRNA AQP9 in RCC clinical specimens. GUSB was used as an internal control. B, C. Spearman’s rank test showed a significant negative correlation between AQP9 expression and miR-532-5p or miR-532-3p levels. D. Immunostaining showed that AQP9 levels were substantially higher in cancer lesions compared to normal tissues with the same staining intensity (100× and 400× magnification field).
Relationship between pre-miR-532 and AQP9 in RCC pathogenesis and clinical outcome
In analyses of TCGA database, patients with high AQP9 expression had significantly advanced tumor stage and pathological grade (Figure 8A-C). A multivariate Cox proportional hazards model showed that high expression of AQP9 was an independent predictive factor for overall survival (hazard ratio [HR]: 1.57, 95% confidence interval [CI]: 1.13-2.17, P = 0.0056) and was similar to well-known clinical prognostic factors such as age, tumor stage and grade (Figure 8D). A combined analysis of overall survival as related to AQP9 and miR-532-duplex expression using the Kaplan-Meier method indicated that high AQP9 and low miR-532-duplex expression levels were significantly associated with poor prognosis compared with low AQP9 and high miR-532-duplex expression levels (P = 0.0025, Figure 8E).
Figure 8.
Clinical significance of AQP9. A-C. High AQP9 expression levels were significantly associated with advanced tumor stage and pathological grade. D. Multivariate analysis for overall survival as a function of clinical parameters, including AQP9 expression. E. Kaplan-Meier plots to analyze overall survival showed that patients with high AQP9 and low miR-532-duplex expression levels had significantly poorer prognosis compared with those that had low and high expression levels of AQP9 and miR-532-duplex, respectively.
Knockdown assay of AQP9 with siRNA
We confirmed that expression levels of both AQP9 mRNA and protein could be suppressed by si-AQP9 transfection of RCC cells (Figure 9A, 9B). Furthermore, AQP9 silencing reduced migration and invasion activity of these cells (Figure 9C-E). A venn diagram of genes that showed significant changes in expression levels following si-AQP9 transfection and in RCC patient samples revealed that 64 genes were downregulated after transfection with si-AQP9 and upregulated in RCC (Figure 9F; Table 4). In a KEGG analysis, we identified five pathways that included genes with differing expression that could be affected by altered AQP9 expression.
Figure 9.
AQP9 knockdown assay by siRNA. A, B. AQP9 mRNA and protein expression 72 h after transfection with si-AQP9_1 or si-AQP9_2 into RCC cell lines. GUSB and GAPDH were used as controls. C-E. Cell proliferation, migration and invasion activities. F. Venn diagram showing downregulated genes after transfection with si-AQP9 and upregulated genes in RCC. KEGG analysis was used to identify enriched pathways. *P < 0.0001, **P < 0.005.
Table 4.
Genes downstream of AQP9 in RCC
| Gene Symbol | Gene Name | Entrez Gene ID | Cytoband | average si-AQP9_1 and 2 transfection in A498 (Log2 ratio) | GEO expression data Fold-Change (Tumor/Normal) | TCGA analysis for OS (high vs low expression: p value) |
|---|---|---|---|---|---|---|
| TLR4 | Toll-like receptor 4 | 7099 | hs|9q33.1 | -2.113 | 1.734 | 0.00171* |
| TIGIT | T cell immunoreceptor with Ig and ITIM domains | 201633 | hs|3q13.31 | -2.074 | 3.185 | 0.0411 |
| CALCRL | Calcitonin receptor-like | 10203 | hs|2q32.1 | -2.025 | 1.826 | 5.31E-06* |
| PABPC4 | Poly (A) binding protein, cytoplasmic 4 (inducible form) | 8761 | hs|1p34.3 | -1.981 | 1.755 | 0.0000124 |
| FOXO6 | Forkhead box O6 | 100132074 | hs|1p34.2 | -1.928 | 1.746 | no data |
| MX2 | MX dynamin-like GTPase 2 | 4600 | hs|21q22.3 | -1.785 | 1.508 | 3.81E-08 |
| DOCK9 | Dedicator of cytokinesis 9 | 23348 | hs|13q32.3 | -1.745 | 1.614 | 1.1E-05* |
| SP140 | SP140 nuclear body protein | 11262 | hs|2q37.1 | -1.707 | 1.800 | 0.000889 |
| CCNE2 | Cyclin E2 | 9134 | hs|8q22.1 | -1.622 | 2.430 | 0.00664 |
| AQP9 | Aquaporin 9 | 366 | hs|15q21.3 | -1.596 | 2.077 | 0.0000203 |
| S100A8 | S100 calcium binding protein A8 | 6279 | hs|1q21.3 | -1.529 | 1.770 | 0.00999 |
| RASA2 | RAS p21 protein activator 2 | 5922 | hs|3q23 | -1.478 | 1.468 | 0.0275* |
| ELK1 | ELK1, member of ETS oncogene family | 2002 | hs|Xp11.23 | -1.433 | 1.514 | 0.0000286 |
| MID1 | Midline 1 | 4281 | hs|Xp22.2 | -1.432 | 1.515 | 0.284 |
| OAS2 | 2’-5’-oligoadenylate synthetase 2, 69/71 kDa | 4939 | hs|12q24.13 | -1.428 | 2.833 | 0.525 |
| CHEK2 | Checkpoint kinase 2 | 11200 | hs|22q12.1 | -1.407 | 1.947 | 2.67E-08 |
| SLC15A3 | Solute carrier family 15 (oligopeptide transporter), member 3 | 51296 | hs|11q12.2 | -1.381 | 4.754 | 0.213 |
| EMILIN2 | Elastin microfibril interfacer 2 | 84034 | hs|18p11.31 | -1.373 | 2.454 | 0.0345 |
| ACTG2 | Actin, gamma 2, smooth muscle, enteric | 72 | hs|2p13.1 | -1.366 | 4.204 | 0.631 |
| BIRC5 | Baculoviral IAP repeat containing 5 | 332 | hs|17q25.3 | -1.355 | 2.729 | 2.93E-09 |
| TIFA | TRAF-interacting protein with forkhead-associated domain | 92610 | hs|4q25 | -1.354 | 1.782 | 0.507 |
| CARD8 | Caspase recruitment domain family, member 8 | 22900 | hs|19q13.33 | -1.293 | 1.650 | 0.854 |
| SNHG1 | Small nucleolar RNA host gene 1 (non-protein coding) | 23642 | hs|11q12.3 | -1.273 | 2.163 | 0.000575 |
| CDK1 | Cyclin-dependent kinase 1 | 983 | hs|10q21.2 | -1.269 | 2.129 | 0.0000667 |
| PLK1 | Polo-like kinase 1 | 5347 | hs|16p12.2 | -1.238 | 2.643 | 3.59E-08 |
| RPL28 | Ribosomal protein L28 | 6158 | hs|19q13.42 | -1.235 | 1.424 | 0.00000308 |
| BTNL9 | Butyrophilin-like 9 | 153579 | hs|5q35.3 | -1.228 | 2.387 | 0.000973* |
| TBC1D10C | TBC1 domain family, member 10C | 374403 | hs|11q13.2 | -1.223 | 3.069 | 0.0116 |
| BDNF | Brain-derived neurotrophic factor | 627 | hs|11p14.1 | -1.216 | 3.163 | 0.0591 |
| GRIA4 | Glutamate receptor, ionotropic, AMPA 4 | 2893 | hs|11q22.3 | -1.214 | 2.292 | 0.0372* |
| SFTA1P | Surfactant associated 1, pseudogene | 207107 | hs|10p14 | -1.207 | 3.020 | 0.869 |
| MIAT | Myocardial infarction associated transcript (non-protein coding) | 440823 | hs|22q12.1 | -1.187 | 1.428 | 5.69E-08 |
| SLC37A1 | Solute carrier family 37 (glucose-6-phosphate transporter), member 1 | 54020 | hs|21q22.3 | -1.172 | 1.512 | 0.0821 |
| RNF149 | Ring finger protein 149 | 284996 | hs|2q11.2 | -1.170 | 2.161 | 0.118 |
| TMEM180 | Transmembrane protein 180 | 79847 | hs|10q24.32 | -1.167 | 2.287 | no data |
| ABL2 | ABL proto-oncogene 2, non-receptor tyrosine kinase | 27 | hs|1q25.2 | -1.160 | 1.431 | 0.237 |
| CXCL13 | Chemokine (C-X-C motif) ligand 13 | 10563 | hs|4q21.1 | -1.160 | 11.383 | 0.00107 |
| ERCC5 | Excision repair cross-complementation group 5 | 2073 | hs|13q33.1 | -1.129 | 1.644 | 0.854 |
| CDKN2A | Cyclin-dependent kinase inhibitor 2A | 1029 | hs|9p21.3 | -1.127 | 3.422 | 0.0103 |
| XAF1 | XIAP associated factor 1 | 54739 | hs|17p13.1 | -1.121 | 2.349 | 0.0000172 |
| ANKRD36BP2 | Ankyrin repeat domain 36B pseudogene 2 | 645784 | hs|2p11.2 | -1.101 | 1.636 | 0.000000142 |
| RDM1 | RAD52 motif containing 1 | 201299 | hs|17q12 | -1.095 | 2.197 | 0.0145 |
| PLK4 | Polo-like kinase 4 | 10733 | hs|4q28.2 | -1.095 | 1.768 | 0.0266 |
| TRIM9 | Tripartite motif containing 9 | 114088 | hs|14q22.1 | -1.091 | 3.764 | 0.552 |
| IFIT2 | Interferon-induced protein with tetratricopeptide repeats 2 | 3433 | hs|10q23.31 | -1.081 | 1.655 | 0.869 |
| TPR | Translocated promoter region, nuclear basket protein | 7175 | hs|1q31.1 | -1.074 | 1.663 | 0.0124* |
| ZNF691 | Zinc finger protein 691 | 51058 | hs|1p34.2 | -1.074 | 1.416 | 0.449 |
| SFMBT2 | Scm-like with four mbt domains 2 | 57713 | hs|10p14 | -1.067 | 2.189 | 0.00977* |
| SLC2A5 | Solute carrier family 2 (facilitated glucose/fructose transporter), member 5 | 6518 | hs|1p36.23 | -1.060 | 1.787 | 0.863 |
| HIST1H3H | Histone cluster 1, H3h | 8357 | hs|6p22.1 | -1.053 | 3.446 | 0.516 |
| MYO1G | Myosin IG | 64005 | hs|7p13 | -1.045 | 2.669 | 0.0204 |
| EPSTI1 | Epithelial stromal interaction 1 (breast) | 94240 | hs|13q14.11 | -1.043 | 1.556 | 0.337 |
| LOXL3 | Lysyl oxidase-like 3 | 84695 | hs|2p13.1 | -1.034 | 1.678 | 0.0012 |
| NR5A2 | Nuclear receptor subfamily 5, group A, member 2 | 2494 | hs|1q32.1 | -1.031 | 2.587 | 0.506 |
| PRR3 | Proline rich 3 | 80742 | hs|6p21.33 | -1.028 | 1.636 | 0.00115 |
| DDX60 | DEAD (Asp-Glu-Ala-Asp) box polypeptide 60 | 55601 | hs|4q32.3 | -1.018 | 1.460 | 0.000848* |
| C1RL | Complement component 1, r subcomponent-like | 51279 | hs|12p13.31 | -1.018 | 1.579 | 0.000000173 |
| TNFRSF4 | Tumor necrosis factor receptor superfamily, member 4 | 7293 | hs|1p36.33 | -1.017 | 2.775 | 0.849 |
| TET3 | Tet methylcytosine dioxygenase 3 | 200424 | hs|2p13.1 | -1.013 | 1.664 | 0.0188 |
| NEURL1B | Neuralized E3 ubiquitin protein ligase 1B | 54492 | hs|5q35.1 | -1.008 | 2.907 | 0.0343* |
| LINGO1 | Leucine rich repeat and Ig domain containing 1 | 84894 | hs|15q24.3 | -1.007 | 2.240 | 0.0344 |
| TNFSF13B | Tumor necrosis factor (ligand) superfamily, member 13b | 10673 | hs|13q33.3 | -1.007 | 4.786 | 0.00527 |
| SQRDL | Sulfide quinone reductase-like (yeast) | 58472 | hs|15q21.1 | -1.003 | 1.899 | 0.784 |
| EVI2A | Ecotropic viral integration site 2A | 2123 | hs|17q11.2 | -1.003 | 2.971 | 0.00737 |
Poor prgonosis in patiets with low gene expression.
Rescue studies by co-transfection of AQP9/miR-532-duplex into A498 cells
We performed AQP9 rescue studies in A498 cells to determine whether oncogenic pathways regulated by AQP9/miR-532-duplex are important for the development of RCC. AQP9 and miR-532-duplex transfection restored AQP9 protein expression (Figure 10A, 10B). Functional assays demonstrated that migration and invasion of RCC cells were significantly recovered by AQP9 and miR-532-duplex transfection compared with cells with restored miR-532-duplex alone (Figure 10C-E).
Figure 10.
Effect of co-transfection of AQP9/miR-532-duplex into A498 cells. A, B. AQP9 protein expression was evaluated 72 h after reverse transfection with miR-532-duplex and 48 h after forward transfection with the AQP9 vector. GAPDH was used as a loading control. C. Cell proliferation assay performed 72 h after reverse transfection with miR-532-duplex and 48 h after forward transfection with the AQP9 vector. D. Cell migration assay performed 48 h after reverse transfection with miR-532-duplex and 24 h after forward transfection with the AQP9 vector. E. Cell invasion assay performed 48 h after reverse transfection with miR-532-duplex and 24 h after forward transfection with AQP9 vector. *P < 0.0001.
Discussion
In contrast to previous theories concerning miRNA biogenesis, we previously revealed that both pre-miRNA strands have functional roles in cancer development (e.g., miR-144, miR-145, miR-455 [9,10,16]. In this study, we demonstrated that expression of both strands of pre-miR-532 (miR-532-5p and miR-532-3p) was suppressed in RCC tissues and that both strands could act as tumor suppressors. Moreover, the oncogenic genes regulated by pre-miR-532 contributed to RCC pathogenesis. Interestingly, previous studies showed that miR-532 has both cancer-promoting and suppressive effects, depending on the type of cancer. For instance, miR-532-5p expression is upregulated in gastric cancer cells and this increased expression was related to an aggressive cancer phenotype [17].
Meanwhile, miR-532-5p expression is downregulated in epithelial ovarian cancer (EOC), in which it functions as a tumor suppressive miRNA that directly targets TWIST1 [18]. Tumor-suppressive functions for miR-532-5p were also reported for other cancer types [19].
In hepatocellular carcinoma (HCC), miR-532-3p (passenger strand) has a tumor-suppressive function by directly regulating the expression of the oncogenic kinesin family member C1 (KIFC1) and the miR-532-3p/KIFC1 axis promoted epithelial to mesenchymal transition (EMT) and HCC metastasis [20]. For RCC, a previous study reported that miR-532-5p expression is downregulated and it functions as a tumor suppressor [21]. The transcription factor ETS1 directly binds the promoter region of pre-miR-532 and negatively controls expression of miR-532-5p in RCC cells [21]. The above results showed that aberrant expression (up- or down-regulation) of miR-532-5p and miR-532-3p is closely associated with cancer pathogenesis. Notably, the two strands of pre-miR-532 have opposite functions depending on the cancer type. Thus, elucidation of molecular mechanisms that regulate the expression of pre-miR-532 is an important issue in cancer research.
For RCC, another area of interest is clarification of which RNA networks are controlled by strands of pre-miR-532. Using our search strategy to identify miRNA targets, in RCC cells we identified 36 and 34 genes as putative oncogenic targets of miR-532-5p and miR-532-3p, respectively. Among these targets, high expression of 19 genes was closely associated with worse prognosis of RCC patients. Analysis of these genes will contribute to a better understanding of the molecular pathology of RCC.
Here we focused on aquaporin 9 (AQP9) because this gene was a putative target for both miR-532-5p and miR-532-3p in RCC cells and its expression was significantly associated with RCC pathogenesis, although its functional significance in RCC is poorly understood. Aquaporin genes (AQPs) encode a family of membrane-spanning water channels, of which 15 are known for mammals [22]. Previous studies showed that some AQPs have aberrant expression in cancer cells that is associated with cancer pathogenesis [23,24]. AQP9 is permeable to small molecules including water, urea, glycerol, arsenite and H2O2, and plays a role in maintaining cellular water homeostasis and energy balance [25]. Previous studies detected overexpression of AQP9 in several cancers such as brain tumor, ovarian cancer and prostate cancer [26-28]. In astrocytoma cells, aberrant AQP9 expression induced AKT activation and decreased E-cadherin expression [29]. Meanwhile, overexpression of AQP9 induced cell cycle arrest and enhanced chemo-sensitivity to 5-fluorouracil in colorectal cancer [30]. The results of these studies suggest that AQP9 can act as an oncogene and thus could be a promising therapeutic target in human cancers.
Our functional assays showed that AQP9 knockdown inhibited cancer cell migration and invasion, and AQP9 overexpression promoted cancer cell aggressiveness in RCC. Aberrant expression of AQP9 was also detected in RCC clinical specimens. We further investigated the expression of AQP9-mediated genes that function downstream of AQP9 in RCC cells. Our data showed that several cell cycle-related genes (e.g., CCNE2, CHEK2, PLK1, CDKN2A and CDK1) were affected by aberrant AQP9 expression. Expression of these cell cycle-related genes was closely associated with RCC patient prognosis. Our data suggested that aberrant expression of AQP9 enhanced RCC oncogenesis and affected several oncogenic genes in RCC cells.
In conclusion, our results showed that expression of both strands of the pre-miR-532-duplex (miR-532-5p and miR-532-3p) was significantly downregulated in RCC clinical specimens and thus the miR-532-duplex could act as an anti-tumour miRNA in RCC. The expression of 19 genes was closely associated with RCC pathogenesis and the expression of these genes was controlled by the pre-miR-532-duplex. Among these targets, AQP9 expression was directly regulated by both miR-532-5p and miR-532-3p in RCC cells. Aberrant expression of AQP9 enhanced cancer aggressiveness, suggesting that AQP9 could be a promising therapeutic target for RCC. Our approach based on antitumor miRNAs could contribute to the development of new diagnostic markers and therapeutic strategies for RCC.
Acknowledgements
The present study was supported by KAKENHI grants 16H05462, 17K11160, 18K09160 and 18K09338, 18K16685, 18K16723 and 18K16724.
Disclosure of conflict of interest
None.
References
- 1.Capitanio U, Montorsi F. Renal cancer. Lancet. 2016;387:894–906. doi: 10.1016/S0140-6736(15)00046-X. [DOI] [PubMed] [Google Scholar]
- 2.Pierorazio PM, Johnson MH, Patel HD, Sozio SM, Sharma R, Iyoha E, Bass EB, Allaf ME. Management of renal masses and localized renal cancer: systematic review and meta-analysis. J Urol. 2016;196:989–999. doi: 10.1016/j.juro.2016.04.081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Figlin R, Sternberg C, Wood CG. Novel agents and approaches for advanced renal cell carcinoma. J Urol. 2012;188:707–715. doi: 10.1016/j.juro.2012.04.108. [DOI] [PubMed] [Google Scholar]
- 4.Bartel DP. MicroRNAs: target recognition and regulatory functions. Cell. 2009;136:215–233. doi: 10.1016/j.cell.2009.01.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Goto Y, Kurozumi A, Nohata N, Kojima S, Matsushita R, Yoshino H, Yamazaki K, Ishida Y, Ichikawa T, Naya Y, Seki N. The microRNA signature of patients with sunitinib failure: regulation of UHRF1 pathways by microRNA-101 in renal cell carcinoma. Oncotarget. 2016;7:59070–59086. doi: 10.18632/oncotarget.10887. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Arai T, Okato A, Kojima S, Idichi T, Koshizuka K, Kurozumi A, Kato M, Yamazaki K, Ishida Y, Naya Y, Ichikawa T, Seki N. Regulation of spindle and kinetochore-associated protein 1 by antitumor miR-10a-5p in renal cell carcinoma. Cancer Sci. 2017;108:2088–2101. doi: 10.1111/cas.13331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kurozumi A, Kato M, Goto Y, Matsushita R, Nishikawa R, Okato A, Fukumoto I, Ichikawa T, Seki N. Regulation of the collagen cross-linking enzymes LOXL2 and PLOD2 by tumor-suppressive microRNA-26a/b in renal cell carcinoma. Int J Oncol. 2016;48:1837–1846. doi: 10.3892/ijo.2016.3440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yamada Y, Arai T, Sugawara S, Okato A, Kato M, Kojima S, Yamazaki K, Naya Y, Ichikawa T, Seki N. Impact of novel oncogenic pathways regulated by antitumor miR-451a in renal cell carcinoma. Cancer Sci. 2018;109:1239–1253. doi: 10.1111/cas.13526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yamada Y, Arai T, Kojima S, Sugawara S, Kato M, Okato A, Yamazaki K, Naya Y, Ichikawa T, Seki N. Anti-tumor roles of both strands of the miR-455 duplex: their targets SKA1 and SKA3 are involved in the pathogenesis of renal cell carcinoma. Oncotarget. 2018;9:26638–26658. doi: 10.18632/oncotarget.25410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Yamada Y, Arai T, Kojima S, Sugawara S, Kato M, Okato A, Yamazaki K, Naya Y, Ichikawa T, Seki N. Regulation of antitumor miR-144-5p targets oncogenes: direct regulation of syndecan-3 and its clinical significance. Cancer Sci. 2018;109:2919–2936. doi: 10.1111/cas.13722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mah SM, Buske C, Humphries RK, Kuchenbauer F. miRNA*: a passenger stranded in RNA-induced silencing complex? Crit Rev Eukaryot Gene Expr. 2010;20:141–148. doi: 10.1615/critreveukargeneexpr.v20.i2.40. [DOI] [PubMed] [Google Scholar]
- 12.Yamada Y, Sugawara S, Arai T, Kojima S, Kato M, Okato A, Yamazaki K, Naya Y, Ichikawa T, Seki N. Molecular pathogenesis of renal cell carcinoma: impact of the anti-tumor miR-29 family on gene regulation. Int J Urol. 2018;25:953–965. doi: 10.1111/iju.13783. [DOI] [PubMed] [Google Scholar]
- 13.Anaya J. OncoLnc: linking TCGA survival data to mRNAs, miRNAs, and lncRNAs. PeerJ Computer Science. 2016;2:e67. [Google Scholar]
- 14.Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, Sun Y, Jacobsen A, Sinha R, Larsson E, Cerami E, Sander C, Schultz N. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal. 2013;6:pl1. doi: 10.1126/scisignal.2004088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, Jacobsen A, Byrne CJ, Heuer ML, Larsson E, Antipin Y, Reva B, Goldberg AP, Sander C, Schultz N. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov. 2012;2:401–404. doi: 10.1158/2159-8290.CD-12-0095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Goto Y, Kurozumi A, Arai T, Nohata N, Kojima S, Okato A, Kato M, Yamazaki K, Ishida Y, Naya Y, Ichikawa T, Seki N. Impact of novel miR-145-3p regulatory networks on survival in patients with castration-resistant prostate cancer. Br J Cancer. 2017;117:409–420. doi: 10.1038/bjc.2017.191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hu S, Zheng Q, Wu H, Wang C, Liu T, Zhou W. miR-532 promoted gastric cancer migration and invasion by targeting NKD1. Life Sci. 2017;177:15–19. doi: 10.1016/j.lfs.2017.03.019. [DOI] [PubMed] [Google Scholar]
- 18.Wei H, Tang QL, Zhang K, Sun JJ, Ding RF. miR-532-5p is a prognostic marker and suppresses cells proliferation and invasion by targeting TWIST1 in epithelial ovarian cancer. Eur Rev Med Pharmacol Sci. 2018;22:5842–5850. doi: 10.26355/eurrev_201809_15911. [DOI] [PubMed] [Google Scholar]
- 19.Song Y, Zhao Y, Ding X, Wang X. microRNA-532 suppresses the PI3K/Akt signaling pathway to inhibit colorectal cancer progression by directly targeting IGF-1R. Am J Cancer Res. 2018;8:435–449. [PMC free article] [PubMed] [Google Scholar] [Retracted]
- 20.Han J, Wang F, Lan Y, Wang J, Nie C, Liang Y, Song R, Zheng T, Pan S, Pei T, Xie C, Yang G, Liu X, Zhu M, Wang Y, Liu Y, Meng F, Cui Y, Zhang B, Liu Y, Meng X, Zhang J, Liu L. KIFC1 regulated by miR-532-3p promotes epithelial-to-mesenchymal transition and metastasis of hepatocellular carcinoma via gankyrin/AKT signaling. Oncogene. 2019;38:406–420. doi: 10.1038/s41388-018-0440-8. [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
- 21.Zhai W, Ma J, Zhu R, Xu C, Zhang J, Chen Y, Chen Z, Gong D, Zheng J, Chen C, Li S, Li B, Huang Y, Xue W, Zheng J. MiR-532-5p suppresses renal cancer cell proliferation by disrupting the ETS1-mediated positive feedback loop with the KRAS-NAP1L1/P-ERK axis. Br J Cancer. 2018;119:591–604. doi: 10.1038/s41416-018-0196-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.De Ieso ML, Yool AJ. Mechanisms of aquaporin-facilitated cancer invasion and metastasis. Front Chem. 2018;6:135. doi: 10.3389/fchem.2018.00135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Nico B, Ribatti D. Aquaporins in tumor growth and angiogenesis. Cancer Lett. 2010;294:135–138. doi: 10.1016/j.canlet.2010.02.005. [DOI] [PubMed] [Google Scholar]
- 24.Papadopoulos MC, Saadoun S. Key roles of aquaporins in tumor biology. Biochim Biophys Acta. 2015;1848:2576–2583. doi: 10.1016/j.bbamem.2014.09.001. [DOI] [PubMed] [Google Scholar]
- 25.Yang JH, Yan CX, Chen XJ, Zhu YS. Expression of aquaglyceroporins in epithelial ovarian tumours and their clinical significance. J Int Med Res. 2011;39:702–711. doi: 10.1177/147323001103900302. [DOI] [PubMed] [Google Scholar]
- 26.Fossdal G, Vik-Mo EO, Sandberg C, Varghese M, Kaarbo M, Telmo E, Langmoen IA, Murrell W. Aqp 9 and brain tumour stem cells. ScientificWorldJournal. 2012;2012:915176. doi: 10.1100/2012/915176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yang JH, Yu YQ, Yan CX. Localisation and expression of aquaporin subtypes in epithelial ovarian tumours. Histol Histopathol. 2011;26:1197–1205. doi: 10.14670/HH-26.1197. [DOI] [PubMed] [Google Scholar]
- 28.Chen Q, Zhu L, Zheng B, Wang J, Song X, Zheng W, Wang L, Yang D, Wang J. Effect of AQP9 expression in androgen-independent prostate cancer cell PC3. Int J Mol Sci. 2016;17 doi: 10.3390/ijms17050738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lv Y, Huang Q, Dai W, Jie Y, Yu G, Fan X, Wu A, Miao Q. AQP9 promotes astrocytoma cell invasion and motility via the AKT pathway. Oncol Lett. 2018;16:6059–6064. doi: 10.3892/ol.2018.9361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Huang D, Feng X, Liu Y, Deng Y, Chen H, Chen D, Fang L, Cai Y, Liu H, Wang L, Wang J, Yang Z. AQP9-induced cell cycle arrest is associated with RAS activation and improves chemotherapy treatment efficacy in colorectal cancer. Cell Death Dis. 2017;8:e2894. doi: 10.1038/cddis.2017.282. [DOI] [PMC free article] [PubMed] [Google Scholar]