The VHL-dependent regulation of microRNAs in renal cancer
© Neal et al. 2010
Received: 21 September 2010
Accepted: 21 October 2010
Published: 21 October 2010
Skip to main content
© Neal et al. 2010
Received: 21 September 2010
Accepted: 21 October 2010
Published: 21 October 2010
The commonest histological type of renal cancer, clear cell renal cell carcinoma (cc RCC), is associated with genetic and epigenetic changes in the von Hippel-Lindau (VHL) tumour suppressor. VHL inactivation leads to induction of hypoxia-inducible factors (HIFs) and a hypoxic pattern of gene expression. Differential levels of specific microRNAs (miRNAs) are observed in several tumours when compared to normal tissue. Given the central role of VHL in renal cancer formation, we examined the VHL-dependent regulation of miRNAs in renal cancer.
VHL-dependent miRNA expression in cc RCC was determined by microarray analysis of renal cell line RCC4 with mutated VHL (RCC4-VHL) and reintroduced wild-type VHL (RCC4 + VHL). Five miRNAs highly upregulated in RCC4 + VHL and five miRNAs highly downregulated in RCC4 + VHL were studied further, in addition to miR-210, which is regulated by the HIF-VHL system. miRNA expression was also measured in 31 cc RCC tumours compared to adjacent normal tissue.
A significant increase in miR-210, miR-155 and miR-21 expression was observed in the tumour tissue. miR-210 levels also showed a correlation with a HIF-regulated mRNA, carbonic anhydrase IX (CAIX), and with VHL mutation or promoter methylation. An inverse correlation was observed between miR-210 expression and patient survival, and a putative target of miR-210, iron-sulfur cluster assembly protein (ISCU1/2), shows reciprocal levels of mRNA expression in the tumours.
We have identified VHL-regulated miRNAs and found that for some the regulation is HIF-dependent and for others it is HIF-independent. This pattern of regulation was also seen in renal cancer tissue for several of these miRNAs (miR-210, miR-155, let-7i and members of the miR-17-92 cluster) when compared with normal tissue. miR-210 showed marked increases in expression in renal cancer and levels correlated with patient survival. The inverse correlation between miR-210 levels and ISCU1/2 provides support for the hypothesis that ISCU1/2 is a target of miR-210 and that it may contribute to the anaerobic respiration seen in renal (and other) tumours.
See Commentary: http://www.biomedcentral.com/1741-7015/8/65
Renal cell carcinoma accounts for 2-3% of malignant diseases in adults, with an increasing worldwide incidence of over 200,000 new cases and 100,000 deaths per year . The understanding of the pathogenesis of the commonest histological type, clear cell renal cell carcinoma (cc RCC), has been considerably advanced by the recognition that the gene encoding the von Hippel-Lindau (VHL) protein is defective in patients with VHL disease and in the majority of sporadic cc RCCs resulting in activation of a hypoxic pattern of gene expression [2, 3]. The exposure of cells to hypoxia normally leads to the coordinated regulation of many genes by the transcription factor hypoxia-inducible factor (HIF). In this process, the VHL protein functions as an E3 ubiquitin ligase which recognises and binds to two specific hydroxyprolyl residues in HIF-1α and HIF-2α and facilitates ubiquitination, leading to rapid proteasomal degradation [4, 5]. In the presence of oxygen, HIF-α molecules undergo prolyl hydroxylation catalysed by three homologous 2-oxoglutarate-dependent dioxygenases, PHD1, PHD2 and PHD3 . The protein products of this broad array of hypoxically regulated genes have critical roles in processes such as energy metabolism, angiogenesis, growth and apoptosis [5, 7, 8]. The majority of cc RCCs exhibit VHL mutations and/or silencing by methylation, which leads to enhanced HIF action and a hypoxic pattern of gene expression [2, 3]. VHL also has a number of other important functions that may have tumour suppressor effects (for review, see ).
Many nonrenal cancers are also characterised by hypoxia, enhanced HIF levels and increased expression of hypoxically regulated genes which correlate both with tumour aggression and patient outcome . Recently, comprehensive gene arrays have emphasised the dominant role of the HIF-VHL transcriptional system and the HIF peptidyl hydroxylases in the regulation of gene expression by hypoxia  and have characterised HIF-dependent and HIF-independent pathways of transcriptional regulation in VHL-deficient cells. However, other mechanisms of gene regulation by hypoxia include control of mRNA stability, regulation of mRNA translation, global influences on the transcriptional machinery and regulation mediated by microRNAs (miRNAs) .
miRNAs are noncoding RNA oligonucleotides that function as important regulators of gene expression. Differential levels of specific miRNAs have been observed in several tumour types when compared to normal tissue [11, 12]. In addition, global reductions in miRNA expression are a feature of many cancers, miRNA gene copy number variation appears common in cancer, and overexpression of miRNAs can contribute to oncogenesis [13–15]. mRNA targets of these miRNAs include genes encoding proteins with roles in apoptosis, cell cycle and growth (for review, see ). Furthermore, certain tumour suppressors such as p53 can directly influence miRNA production , and gene polymorphisms of components of the miRNA biogenesis machinery have been associated with renal cell carcinoma susceptibility . The precise mechanisms underlying inducible production of miRNAs, the role of tumour suppressors, such as VHL, in miRNA regulation and the effects of the HIF-VHL system on miRNA activity require further definition, as does the contribution of VHL-dependent alterations in miRNA abundance to the pathogenesis of renal cancers.
To determine the possible role of miRNAs in hypoxic gene regulation and to examine for hypoxically regulated miRNAs that may have relevance to tumour pathogenesis, we have surveyed changes in miRNA expression levels in response of breast cancer cells to hypoxia and characterised the hypoxic regulation of one specific human miRNA (miR-210) [19, 20]. Given the link between hypoxia and miR-210 regulation and the recognition of particular alterations in miRNA expression in cancer, we hypothesised that levels of miR-210 expression in cancer would correlate with degree of hypoxia and tumour behaviour. To test this hypothesis, we have recently examined the expression of this hypoxically induced miRNA in human breast cancers and found a striking association with breast cancer prognosis [19, 20]. Given the central role of VHL in hypoxic gene regulation and in renal cancer formation, we wished to examine the VHL-dependent regulation of miRNAs in renal cancer. miRNA dysregulation has been observed in renal cancer, and miR-210 has been identified as a hypoxia and VHL-regulated miRNA in renal cancer cells [21–23]. We wished to understand the mechanism underlying such alterations, examine the extent to which VHL-mediated alterations in miRNAs were HIF-dependent or independent and examine for the utility of miRNA alterations in renal cancer diagnosis.
The renal cancer cell line RCC4 stably transfected with either an empty vector or a vector encoding VHL was used in this study . Cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Invitrogen, Newcastle, NSW, Australia). All experiments were conducted in triplicate with independent cell cultures.
Treatment of cells with dimethyloxalylglycine (DMOG) (Biomol International, Plymouth Meeting, PA, USA) involved supplementing cell media with 1 mM DMOG diluted in phosphate-buffered saline (PBS) for 24 hours [25, 26].
For HIF-1α and HIF-2α small interfering RNA (siRNA) treatments, RCC4-VHL cells were seeded at 30-50% confluency and grown for 24 hours. Cells were then transfected with 50 mM HIF-1α (sense 5'-CUGAUGACCAGCAACUUGAdTdT-3' and antisense 5'-UCAAGUUGCUGGUCAUCAGdTdT-3'; Dharmacon, Lafayette, CO, USA), HIF-2α siRNA (sense 5'-CAGCAUCUUUGAUAGCAGUdTdT-3' and antisense 5'-ACUGCUAUCAAAGAUGCUGdTdT-3'; Dharmacon)  and Accell negative control kit (Dharmacon) using Lipofectamine 2000 (Invitrogen). After 48 hours, RNA and protein were extracted.
For miR-210 overexpression studies, 20 nM miR-210 oligonucleotide duplex (sense 5'-CUGUGCGUGUGACAGCGGCUGA-3' and antisense 5'-AGCCGCUGUCACACGCACAGUU-3'; GenePharma, Shanghai, China) and 20 nM mimic negative control (sense 5'-UUCUCCGAACGUGUCACGUTT-3' and antisense 5'-ACGUGACACGUUCGGAGAATT-3'; GenePharma) were transfected into RCC4 + VHL cells seeded at 30-50% confluency. RNA was harvested after 48 hours.
For miR-210 repression studies, 20 nM miR-210 siRNA (5'-UCAGCCGCUGUCACACGCACAG-3'; GenePharma) and 20 nM miRNA inhibitor negative control (5'-UCUACUCUUUCUAGGUUGUGA-3'; GenePharma) were transfected into RCC4-VHL cells seeded at 30-50% confluency. RNA was harvested after 48 hours.
The miRNA microarrays consisted of 1488 antisense miRNA oligonucleotide probes (miRCURY LNA microRNA probe set, catalog no. 208010 V8.1; Exiqon, Vadbaek, Denmark) printed in duplicate onto epoxide-coated microarray slides (Corning Life Sciences, Acton, MA, USA). For detection on the array, 5 μg of total RNA was labelled by the ligation of a fluorescently modified RNA dimer . Two sample (dual colour) competitive hybridizations were performed using Cy3- and Cy5-labelled sample pairs.
Hybridisation was performed for 16 hours at 56°C under LifterSlips (Erie Scientific, Portsmouth, NH, USA) in 1× Exiqon hybridization buffer (catalog no. 208020) (Exiqon, Vadbaek, Denmark) in a total volume of 25 μL.
Slides were placed in Corning hybridization chambers and protected from light for the 16-hour incubation. Slides were washed using dilutions of the Exiqon Wash Buffer kit (catalog no. 208021) as recommended by the manufacturer. Slides were scanned at 10-μm resolution with a Genepix 4000B Scanner (Molecular Devices, Union City, CA, USA).
Median spot pixel intensity values in scanned images were extracted using the Spot v3 plugin (CSIRO, Clayton South, VIC, Australia) for the statistical environment R. After subtraction of background intensities and global loess normalisation, mean intensities were log2 transformed and ratios (Cy5/Cy3) were obtained. Differentially expressed miRNAs were determined using linear models and empirical Bayesian moderation of standard errors (LIMMA R package, WEHI, Melbourne, VIC, Australia; ).
Whole cell extracts were resolved by standard polyacrylamide gel electrophoresis and electroblotted onto polyvinylidene difluoride membrane (Millipore, Bedford, MA, USA). Primary antibodies used were mouse monoclonal anti-HIF-1α (610958; BD Transduction Laboratories, San Diego, CA, USA) and rabbit polyclonal anti-HIF-2α (NB100-122; Novus Biologicals, Littleton, CO, USA). Horseradish peroxidase-conjugated secondary antibodies goat anti-rabbit IgG and donkey anti-mouse IgG (Immunopure; Thermo Scientific, Rockford, IL, USA) were used in conjunction with the ECL system (SuperSignal West Pico; Pierce, Rockford, IL, USA) to visualise bands using an ImageQuant LAS 4000 system (GE Healthcare Life Sciences, Uppsala, Sweden).
Summary of the clinical characteristics of 31 patients with cc RCC.
Number of patients
miRNA expression was assessed by relative quantitation real-time PCR (qPCR) using TaqMan microRNA assays (Applied Biosystems, Foster City, CA, USA).RNA was extracted from cells and tissue using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. RNA quantity and quality was determined using a Nanodrop-8000 spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA). cDNA was synthesised from 5 ng of total RNA using TaqMan miRNA-specific primers and the TaqMan microRNA reverse transcription kit (Applied Biosystems). qPCR was performed using the Corbett Rotor-gene 2000. Each PCR was performed in triplicate and contained 1 μl of reverse transcription product, 1× TaqMan Universal PCR Master Mix No AmpErase UNG and 0.5 μl of primer and hydrolysis probe mix of the TaqMan microRNA assay (assay IDs: miR-210: 000512, miR-155: 000479, miR-21: 000397, miR-31: 001100, miR-20a: 000580, miR-18a: 002422, miR-17: 000393, let-7i: 002221, miR-193b: 001010; Applied Biosystems). The 10-μl reactions were incubated at 95°C for 10 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 60 seconds. Results were normalised to the expression of the small nuclear RNA gene RNU6B (assay ID: 001093; Applied Biosystems) for cell samples and an average of the small nucleolar RNAs RNU43 and RNU48 (assay IDs: 001095 and 001006, respectively; Applied Biosystems) expression levels for tissue samples, as no significant variation of expression was found between tumour and normal tissue. Data were generated and analysed using Corbett Rotorgene software (version 5.0.61) (Corbett Research, Sydney, NSW, Australia) and the Relative Expression Software Tool (REST) program (Corbett Research, Sydney, NSW, Australia) .
For mRNA expression studies, 1 μg of RNA was reverse transcribed following DNase treatment (New England Biolabs, Beverly, MA, USA) using M-MLV Reverse Transcriptase RNase H minus, Point mutant (Promega, Madison, WI, USA) and Oligo dT(15) primers. Carbonic anhydrase IX (CAIX) mRNA expression was assessed by qPCR according to the SYBR Green protocol (Applied Biosystems). The following primers were used to amplify CAIX (each at 0.25 μM per reaction): forward 5'-CCTCTCCCGGAACTGAGCCTAT-3' and reverse 5'-TGTTCTGAGCCTGGGTGATCTG-3' . Iron-sulfur cluster assembly (ISCU1/2) mRNA expression was assessed by qPCR using the ISCU1/2 TaqMan gene expression kit (assay ID: Hs00384510_m1; Applied Biosystems) according to the manufacturer's instructions, and the human β-actin gene was used as a reference using the following primers (each at 0.25 μM per reaction): forward 5'-TTGCCGACAGGATGCAGAAG-3' and reverse 5'-GCCGATCCACACGGAGTACT-3'.
Mutation analysis of the promoter region, three exons and associated splice junctions of the VHL gene were performed by PCR amplification and cycle sequencing. Genomic DNA was extracted from tumour samples using the DNeasy kit, spin-column protocol (Qiagen, Hilden, Germany) according to the manufacturer's instructions. DNA was PCR amplified in 50-μl reactions using AmpliTaqGold (Roche, Branchburg, NJ, USA)according to the manufacturer's instructions and the following conditions and primers. Promoter and Exon 1: Primers forward 5'-TAGCCTCGCCTCCGTTACA-3' and reverse 5'-GCTTCAGACCGTGCTATCG-3', Exon 2: Primers forward 5'-TGATCTCCTGACCTCATGAT-3' and reverse 5'- GACACCATAACACCTTTAAC-3' and Exon 3: Primers forward 5'-TACTGAGACCCTAGTCTG-3' and reverse 5'-GGAAGGAACCAGTCCTG-3'. PCR primers for exons 1 and 3 were tagged with universal M13 primers. PCR products were visualised on agarose gel and purified using ExoSAP-IT (GE Life Sciences) according to the manufacturer's instructions. Cycle sequencing was performed using universal M13 forward and reverse primers for exons 1 and 3 and the PCR primers for exon 2 and standard BigDye (Applied Biosystems) chemistry. Sequencing products were purified using Agencourt CleanSEQ Dye Terminator Removal (Beckman Coulter, Beverley, MA, USA) according to the manufacturer's instructions and analysed on an Applied Biosystems 3730 analyser.
Sequence traces were compared to the National Center for Biotechnology Information (NCBI) reference sequence NM_000551.2 using Mutation Surveyor (SoftGenetics, State College, PA, USA). Splice site variants were analysed using the Berkeley Drosophila Genome Project (BDGP) http://www.fruitfly.org/about/index.html, NetGene2 http://www.cbs.dtu.dk/services/NetGene2/ and GeneSplicer http://www.cbcb.umd.edu/software/GeneSplicer/gene_spl.shtml splice site predictors. Human Genome Variation Society (HGVS) nomenclature was used for variant identification .
Bisulphite treatment followed by methylation-specific PCR (MSP) was used to analyse the methylation status of the VHL promoter . Genomic DNA was subjected to bisulphite treatment using an EpiTect Bisulfite kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The following primers were used for amplifying the VHL promoter: Unmethylated-specific (expected size 165 bp) forward 5'-GTTGGAGGATTTTTTTGTGTATGT-3', reverse 5'-CCCAAACCAAACACCACAAA-3', and methylated-specific (expected size 158 bp) forward 5'-TGGAGGATTTTTTTGCGTACGC-3', reverse 5'-GAACCGAACGCCGCGAA-3' . The PCR mixture contained 1× PCR buffer (New England Biolabs), deoxynucleotide triphosphates (each at 2.5 mM), primers (each at 0.8 mM), 50 ng modified DNA and 2.5 U of Taq DNA polymerase (New England Biolabs) added after a 5-minute at 95°C hot start, in a total reaction volume of 50 μl. Reactions were carried out on a GeneAmp PCR system 9700 thermocycler (Applied Biosystems) for 35 cycles (30 seconds at 95°C, 30 seconds at 47°C and 30 seconds at 72°C), followed by a final extension of 4 minutes at 72°C. Each PCR was loaded (20 μl) onto a 3% agarose gel containing ethidium bromide and visualised under UV light.
All statistical analysis was done using PASW Statistics 17 (Somers, NY, USA). The Mann-Whitney U test was used to assess the difference between miRNA expression in RCC4 ± VHL cells, treated/untreated cells and tumour/normal adjacent tissue. Spearman's ρ correlation coefficient was used to assess correlations between miRNA expression and ISCU1/2 and CAIX expression and for the correlation of miR-210, miR-155, miR-21, miR-18a, let-7i, ISCU1/2 and CAIX expression with survival, tumour size, grade and tumour node metastasis (TNM) stage. Kaplan-Meier log-rank (Mantel-Cox) analysis was undertaken to assess overall survival, miR-210 expression and TNM stage. Results were considered statistically significant at P ≤ 0.05.
Given these interesting observations in one particular renal cancer cell line, we examined whether this signature of VHL-dependent miRNA dysregulation occurs in renal tumours. This would provide information about the extent to which these specific VHL-dependent miRNA alterations are more widely observed in renal cancer, with important implications for understanding renal oncogenesis and developing diagnostic and prognostic markers in renal cancer.
These results show that some of the VHL-regulated miRNAs, such as miR-210 and miR-155 in cultured cells, are highly differentially expressed between renal cancer tissue and normal tissue. However, other miRNAs such as miR-31 and miR-193b do not show such regulation and indeed, somewhat surprisingly, miR-21 and let-7i, which were decreased in cells lacking VHL, were overexpressed in renal cancer when compared to normal adjacent renal tissue. This suggests that despite VHL regulation at the cellular level, tumour growth and progression select for enhanced miR-21 and let-7i expression in tumour cells or other cells within the renal cancers. The decrease in expression in RCC4-VHL cells may be an artefact of cell culture resulting from different selection pressures in culture compared to cc RCC tumour tissue. Also, there are many mechanisms that may lead to overexpression of miR-21 in renal cancer, and these may not be related to VHL function. The RCC4 cell line used in this study may not be representative of the situation in all renal cancers, and in future studies it would be beneficial to validate these findings in additional kidney cancer cell lines.
miR-21 has been identified as having oncogenic properties and is overexpressed in the majority of cancer types examined . A host of tumour suppressor genes have now been validated as targets of miR-21 (for review, see ), and recently the suggestion has been made, on the basis of bioinformatics analysis, that VHL may be a target of miR-21 [22, 37], but this awaits experimental validation. Our finding that miR-21 is upregulated in cc RCC tumours is in keeping with two recent studies profiling miRNA expression in renal cancer that have also identified miR-21 as upregulated in cc RCC tumour tissue compared to normal kidney tissue, although the increase in expression observed in our study (fourfold) was higher than previously reported (2.5- and 1.2-fold) [21, 22].
The overexpression of miR-210 in cc RCC has also recently been reported [21–23]. miR-210 is also overexpressed in many other tumour types , most notably breast cancer, in which it was found to be a prognostic marker . miR-155 is also overexpressed in many cancers  and has been shown to inhibit apoptosis by targeting tumour protein p53 inducible nuclear protein 1 (TP53INP1) . In addition, O'Connell and others have found that inositol phosphatase SHIP1 is a target of miR-155 . Our results are in accord with those of other studies which have found increased miR-155 expression levels in cc RCC tissue, although the reported fold increases (3.2- and 6.4-fold) [21, 23] were less than what was found in this series (15-fold).
The differential expression of some members of the miR-17-92 cluster in cc RCC is particularly interesting, as overexpression of this miRNA cluster has been implicated in a wide range of cancers and has been shown to act as an oncogene [13–15]. miR-17-92 has been shown to play an important role in tumour cell proliferation and apoptosis [16, 40], to negatively regulate HIF-1α expression  and appears to have a critical function in vascular endothelial growth factor-induced angiogenesis . Under normal physiological conditions, miR-17-92 is involved in the regulation of MYC-induced cell proliferation by inhibiting E2F1 expression; however, when miR-17-92 is overexpressed in many cancers, it can act with MYC to synergistically contribute to aggressive cancer development [14, 43–45]. The overexpression of three members of the miR-17-92 cluster, miR-17, miR-20a and miR-18a, has previously been reported for cc RCC [21, 37]. Our finding that these three miRNAs are upregulated in the RCC4-VHL cell line is in accord with findings in these studies; however, we were unable to confirm the upregulation of miR-17 in tumour tissue or the statistically significant upregulation of miR-20a. This discrepancy may be due to small sample size, differences in procedure or potential interference of other cell types in the tumour samples.
Eleven members of the let-7 family have been identified in the human genome. The family was one of the earliest tumour suppressor miRNAs identified in cancer, and some members of the let-7 family can also have oncogenic roles . The increased expression of one member of the let-7 family, let-7i, has been reported in breast cancer  and head and neck squamous cell carcinoma , and decreased expression is associated with chemotherapy resistance and shorter progression-free survival in late-stage ovarian cancer patients . To the best of our knowledge, this is the first report of a change in let-7i expression in renal cancer.
However, a substantial proportion of the tumours without identified VHL inactivation still display high expression of CAIX and miR-210. Whilst it is possible that our mutational and methylation analysis did not identify all of the causes of VHL inactivation present in these tumours, it indicates that they may have other genetic mutations that influence VHL or HIF expression or function. These mechanisms potentially include disruptions in SETD2 (histone H3K36 methyltransferase) or JARID1C (histone H3K4 demethylase), which have both been recently identified as having truncating mutations in approximately 3% of cc RCC tumours . Other potential mechanisms of VHL suppression might be mediated via miRNAs, as has been reported recently for miR-92 in chronic lymphocytic leukaemia . The mechanisms leading to cc RCC tumour initiation and/or activation of hypoxic pathways in tumours without VHL mutation or VHL promoter methylation are not well understood, and further work is needed.
A significant correlation was also found between miR-155 and miR-21 expression in cc RCC tumours and tumour size (mm) (r = 0.408, P = 0.028; r = 0.434, P = 0.019, respectively) shown in Figures 8c and 8d. Correlations were also observed for miR-210 and let-7i expression and tumour size (mm); however, these correlations did not achieve statistical significance (r = 0.356, P = 0.058; and r = 0.363, P = 0.053, respectively). No correlations were found between miRNA expression in cc RCC tumours and tumour grade or TNM stage.
The relationship between miR-210 levels and patient outcome suggests that miR-210 levels in cancer are not simply a marker of hypoxia but have a fundamental influence on tumour behaviour. To understand the mechanisms for the effects of miR-210 levels on patient outcome, we wished to study influences on potential mRNA targets. Computer-based algorithms are capable of predicting a large number of potential mRNA targets of particular miRNAs, some of which have been validated experimentally. We have defined a large number of potential miR-210 targets using such methodology, and during the course of this study Chan and colleagues  and subsequently others [58–60] were able to experimentally validate the regulation of one such predicted target, ISCU1/2.
The iron cluster assembly proteins, ISCU1 and ISCU2, are involved in the biogenesis of [4Fe-4S] and [2Fe-2S] iron-sulfur clusters, which are implicated in a wide range of biological processes, including electron transport and mitochondrial oxidation-reduction reactions . The ISCU protein exists as two splice isoforms in mammalian cells which share both structural and functional similarities: ISCU1 is located in the cytosol, whereas ISCU2 is located in the mitochondria . It has been hypothesised that miR-210 represses the expression of ISCU1/2 during hypoxic conditions, thus disrupting mitochondrial function and causing downstream metabolic changes within the cell [57–60].
We sought to examine whether the elevation of miR-210 levels in cc RCC was associated with reduced expression of ISCU1/2 mRNA. The expression of ISCU1/2 was analysed in the RCC4 ± VHL cell lines, and it was found that the expression of ISCU1/2 was significantly reduced in the RCC4-VHL line by at least fivefold (P = 0.05) (Additional file 4), in keeping with targeting of ISCU1/2 by increased miR-210 expression. Consistent with this finding, a miR-210 mimic transfected into RCC4 + VHL cells to increase miR-210 levels did markedly decrease the mRNA expression of ISCU1/2 (threefold; P = 0.05) (Additional file 5a). Likewise, a miR-210 antagomir transfected into RCC4-VHL cells to decrease miR-210 levels markedly increased the mRNA expression of ISCU1/2 (twofold; P = 0.05) (Additional file 5b).
In this study, we have identified miRNAs which are regulated by VHL in cell culture. For some of these miRNAs, the regulation was HIF-dependent and for others it was HIF-independent. This pattern of regulation was also seen in cc RCC for some of these miRNAs when compared with normal adjacent renal tissue (miR-210, miR-155, let-7i and miR-18a). The level of one HIF- and VHL-regulated miRNA, miR-210, showed marked increases in expression in the renal cancer tissue, and expression levels were correlated inversely with patient survival. The enhanced levels of miR-210 in renal cancer are likely to be mediated by unrestrained HIF activity, independent of hypoxia. It suggests that the association that has been seen in other cancers between miR-210 levels and patient survival is not solely because it is a marker of tumour hypoxia, but because it influences tumour behaviour through alterations in gene expression. miR-210 levels also showed a correlation with a HIF-regulated mRNA, CAIX, and with the presence of VHL mutation or promoter methylation. However, some tumours without evidence of VHL inactivation also had elevated miR-210 and CAIX levels, indicating the likely operation of other mechanisms of HIF activation. We also found a strong inverse correlation between miR-210 levels and mRNA expression of a miR-210 target gene, ISCU1/2, which may contribute to the repression of mitochondrial proteins by VHL and to the anaerobic pattern of respiration seen in renal (and other) tumours.
Carbonic anhydrase IX
clear cell renal cell carcinoma
Iron-sulfur cluster assembly protein
methylation specific PCR
relative quantitation real-time PCR
renal cell carcinoma
small interfering RNA
Tumour Node Metastasis
This research was supported by a Research Establishment Award from the Don and Lorraine Jacquot Foundation administered by the Royal Australasian College of Physicians. The authors thank Virginia Papangelis at the Flinders Medical Centre/Repatriation General Hospital tissue bank facility; Dr Richard Hummel, Department of Surgery, Flinders Medical Centre, for help with statistical analysis; Dr Ben Roberts, Division of Health Sciences, University of South Australia, for providing the RCC4-/+ VHL cell lines; and Belinda Mercorella, Department of Genetics and Molecular Pathology, IMVS, for assistance with VHL gene sequencing.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.