Background
Hypoxia is a key feature of many cancers and the presence of hypoxia is associated with more aggressive and metastatic tumours [
1‐
3]. The exposure of cells to hypoxia leads to the co-ordinated regulation of many genes. The protein products of these genes have a wide variety of critical roles in processes such as metabolism, angiogenesis, growth and apoptosis. Studies of the mechanisms underlying the regulation of such genes have implicated a central role for the transcription factor hypoxia inducible factor (HIF). Many cancers are characterised by enhanced HIF levels and increased expression of hypoxically regulated genes which correlate both with tumour aggression and patient outcome [
4]. The extent to which hypoxia contributes to enhanced tumour aggression via metabolic alterations, changes in gene expression and/or epigenetic modifications remains incompletely understood. In addition to the transcriptional regulation of mRNA, hypoxia can also influence mRNA stability, translation, protein stability and microRNA generation. Hypoxia also leads to the co-ordinated repression of many genes but the mechanism for such effects is less well defined.
MicroRNAs (miRNAs) are 17–22 nucleotide, non-coding, single stranded RNA molecules that are important regulators of gene expression. MicroRNAs are transcribed as primary transcripts (pri-miRNAs) by RNA Polymerase II [
5] or III [
6] enzymes. In the nucleus, pri-miRNAs are cleaved by the nuclear RNase III enzyme Drosha and cofactor protein DGCR8 (DiGeorge syndrome critical region gene 8) to generate a precursor miRNA (pre-miRNA) (about 70 nucleotides long) [
7]. The pre-miRNAs are exported to the cytoplasm by the karyopherin, exportin-5 [
8]. In the cytoplasm, pre-miRNAs are further cleaved by Dicer, a ribonuclease III enzyme [
9,
10], coupled with TARBP2 (Tar RNA binding protein 2) [
11,
12] to generate a 22 nucleotide double stranded miRNA duplex. One strand of the RNA duplex functions as the mature miRNA and often the passenger strand will be degraded by endonucleases [
13]. In some instances the passenger strand (previously referred to as the star form) is not degraded and may also function as a mature miRNA [
14,
15]. The two strands of the duplex are separated by an RNA helicase (DDX5) and the miRNA then binds to mRNA within the RNA induced silencing complex, containing an Argonaute protein, leading to either translational inhibition or destruction of the target mRNA [
16].
Functional studies show that miRNAs are involved in regulating many cellular processes including developmental timing, cell differentiation, cell proliferation and cell death [
17,
18]. The expression levels of many miRNAs are deregulated in human disease conditions including cancer [
19‐
23]. In addition to deregulation of specific miRNAs in cancer, it has emerged that most tumour cell lines and cancers are characterised by global reductions in miRNA expression [
24,
25] when compared to adjacent normal tissue. Whilst it has been postulated that this reduction in miRNA expression is a feature of a loss of differentiation, the mechanisms underlying this global reduction in miRNA expression in many cancers are unknown, though some evidence exists for post-transcriptional control [
25,
26].
It has recently been suggested that genetic polymorphisms of the genes encoding miRNA generating proteins may affect renal cell carcinoma susceptibility [
27]. Indeed for other cancers such as ovarian and lung tumours, low levels of Drosha [
28] and Dicer [
28,
29] have been shown to correlate with poor patient outcome. Dicer1 is a haploinsufficient tumour suppressor in human cancer and loss of one Dicer1 allele is sufficient for the formation of tumours in breast, kidney, stomach, intestine, liver, lungs and pancreas [
30]. Furthermore, miRNAs are central to the process of angiogenesis (for review see [
31]). Exposure to hypoxia alters specific miRNA expression [
32‐
35] with miRNAs such as miR-210 showing marked hypoxic induction and capacity to act as markers of patient prognosis in breast cancer [
35].
The links between tumour hypoxia, miRNA expression and cancer aggression raise the possibility of a general effect of hypoxia on miRNA biogenesis and function. During our microarray study examining mRNA expression in the breast cancer cell line MCF7, we saw a modest but consistent decrease in Dicer mRNA levels after exposure to hypoxia [
36]. In this work we investigate the possibility that hypoxic regulation of expression of miRNA biogenesis proteins might contribute to the reduction in miRNA expression in many tumours and to the role of hypoxia in cancer progression.
Methods
Cell culture
Breast cancer cell lines (MCF7 and SKBR3) and colorectal cancer cell line (HT29) were obtained from the American Type Culture Collection (Manassas, VA, USA). The collection of primary human umbilical vein endothelial cells (HUVEC) for use in this study was given ethical clearance from the Royal Adelaide Hospital (RAH), Adelaide, South Australia. Consent was obtained from all subjects in accordance with the ‘Declaration of Helsinki’ and conforms to the guidelines established by the National Health and Medical Research Council of Australia.
Breast cancer cell lines (MCF7 and SKBR3) and colorectal cancer cell line (HT29) were maintained in RPMI 1640 (Invitrogen) medium supplemented with 10% foetal bovine serum (FBS) (Bovigen). Primary cell line human umbilical vein endothelial cells (HUVECs) were maintained in polystyrene flasks coated with gelatine in Media 200 PRF (Invitrogen) supplemented with 20% FBS (Bovigen). Cells were maintained at 37°C with 5% CO2. All experiments were conducted in triplicate with independent cell cultures. Independently treated cultures were treated as technical replicates.
Cell treatments
Cells were treated with 1 mM dimethyloxalylglycine (DMOG) (Enzo Life Sciences) and incubated at 37°C in a normoxic incubator for 48 h. Similarly cells were treated with 0.1 mM Desferrioxamine (Sigma-Aldrich) and incubated at 37°C in a normoxic incubator for 48 h.
Exposure of cells to hypoxia
A hypoxic incubator (Coy Laboratory Hypoxic workstation glove box) was used to expose cells to continuous controlled hypoxic conditions. This humidified, temperature controlled (37°C) chamber was supplemented with 5% CO2, and N2 (as required to maintain controlled O2 levels). Cells were exposed to 0.1% O2 levels or 1% O2 for varying durations. Normoxic controls were incubated in parallel in a humidified incubator supplemented with 5% CO2 at 37°C.
RNA interference
Cells were seeded at 5 × 10
4 cells per well in 24-well plates (1.9 cm
2) and grown for 24 h. Cells were transfected with 20 nM siRNA duplexes (Shanghai GenePharma Co., Ltd, China), using Lipofectamine 2000 reagent (Invitrogen) following the manufacturer’s protocol. A second transfection was carried out after 24 h following the same protocol. Cells were harvested 24 h after the second transfection and used for RNA and protein extraction. List of siRNAs used are provided (see Additional file
1: Table S1).
Transfection with miRNA mimics and inhibitors
Cells were seeded at 5 × 104 cells per well in 24-well plates and grown for 24 h. Cells were transfected with 20 nM miRNA mimic, inhibitors or antagomirs (Shanghai GenePharma Co., Ltd, China), using Lipofectamine 2000 reagent (Invitrogen) following the manufacturer’s protocol. Cells were harvested 24 h after the transfection and used for RNA and protein extraction.
Plasmid DNA transfection
A plasmid construct containing the 3’UTR of the ZEB1 gene, downstream of
Renilla luciferase (RL), in a CMV promoted RL reporter (pCI-neo-hRL) was used [
37]. ZEB1 is an E-cadherin transcriptional repressor involved in epithelial to mesenchymal transition of tissues and is regulated by the mir-200 family. To over-express miR-200b levels we used a plasmid expressing a precursor miR-200b from a CMV promoter (pCMV-miR-200b) and an empty vector was used as the control plasmid (Origene).
The RL reporter plasmids (3.6 fmol), pGL3-control (Promega) (500 ng for normalisation) and pCMV-miR-200b/pCMV-miR plasmid (250 ng) were co-transfected with Lipofectamine 2000 (Invitrogen) into SKBR3 cells seeded in 24-well plates (1 × 105 cells per well). The total amount of DNA in each transfection was made up to 1 μg with unrelated plasmid DNA (pcDNA 3.1+). Two hours after transfection, half the plates were incubated at 0.1% O2 for 24 h and control plates were incubated at normoxic conditions. After the 48 h incubation cells were assayed using the dual-luciferase reporter assay system (Promega). All experiments were performed in triplicate. Luminescence was measured using a plate reader luminometer (Beckman Coulter DTX 880 Multimode detector).
RNA extraction and real time PCR
Total RNA was extracted using TRIzol reagent (Invitrogen) following the manufacturers protocol. RNA quantity and quality were determined using a Nanodrop-8000 spectrophotometer (Nanodrop Technology) and Agilent 2100 Bioanalyzer. For miRNA analysis complementary DNA (cDNA) was synthesised from 5 ng of total RNA using Taqman miRNA specific primers and Taqman miRNA reverse transcription kit (Applied Biosystems). Small nucleolar RNA, RNU6B, was used as a control gene. For parallel detection of mature and pre-miRNAs, cDNA was synthesised from 1 μg of RNA using the miScript II RT kit. Mature miRNAs were analysed using miScript primer assays and precursor miRNAs were analysed using miScript precursor assays. For mRNA analysis cDNA was randomly primed from 1 μg of total RNA following DNase 1 treatment (New England Labs) using M-MLV reverse transcriptase RNase H minus, point mutant (Promega) and Random primer 6 (New England Labs). Real time PCR was subsequently performed in triplicate with 1:5 dilution of cDNA using Taqman gene expression assays. Beta-2-microglobulin and ribosomal RNA 18S were used to normalise mRNA expression. Relative quantification by RT-PCR was performed using the Corbett Roto-gene 6000 and data analysed using Corbett Rotogene software (Version 5.0.61) (Corbett Research).
Microarray analysis
Total RNA from MCF7 cells, exposed to hypoxia or normoxia, was extracted using the TRIzol protocol as above. RNA integrity was assessed using the Agilent 2100 Bioanalyzer. Affymetrix miRNA 3.1 Array Strip was used for RNA analysis. This array consisted probe sets unique to human mature and pre-miRNA hairpins. A detailed protocol can be found in the miRNA 3.1 Array Strips technical manual (Affymetrix). In summary, 100–300 ng of total RNA was used to synthesise double stranded cDNA using random hexamers. The cDNA was then amplified to produce antisense cRNA, which was then reverse transcribed in a second cycle of cDNA synthesis. The second cycle incorporates dUTP into the cDNA sequence, which allows it to be fragmented using uracil DNA glycosylase and apurinic/apyrimidic endonuclease I. Following biotinylation, these fragments were hybridised overnight to a Affymetrix miRNA 3.1 array. The arrays were then washed, stained using a fluorescently-labelled antibody, and scanned using a high-resolution scanner. Intensity data were analysed using Partek® software (Partek Inc.). Data were normalised by quantile normalisation and log
2 transformed. Differential expression was determined by ANOVA and corrected for false discovery. The microarray data have been deposited in NCBI’s Gene Expression Omnibus [
38] and are accessible through GEO Series accession number GSE49999.
Immunoblotting
Cell lysates were prepared by adding lysis buffer (6.7 M urea, 10 mM Tris–HCl (pH 6.8), 10% glycerol, 1% SDS, supplemented with 1 mM DTT and Complete mini protease inhibitor cocktail tablets (Roche Applied Science, UK) just before use, to cells washed with 1× phosphate buffered saline solution. Total protein extracts were resolved on an 8% SDS polyacrylamide gel or a Mini-PROTEAN TGX precast Any kD gradient gel (Bio-Rad Laboratories, Inc.) electrophoresis and transferred to a polyvinylidene difluoride membrane (Millipore). The use of Mini-PROTEAN TGX precast gels allowed for imaging the total protein loaded in each lane on the gel and on the PVDF membrane after the transfer. The membrane was blocked with 5% skim milk in 1xTBS-T for 1 h at room temperature and incubated with primary antibody overnight at 4°C, then washed with 1xTBS-T and incubated with the appropriate horseradish peroxidise conjugated secondary antibodies for 1 h at room temperature. The enhanced chemiluminescence (ECL) (GE health care) system was used to detect bands, which were imaged with a Chemi-Doc MP system (Bio-Rad). Densitometry was performed using the ImageLab software version 4.0 (BioRad). Protein levels were normalised to β-tubulin, α-actinin or to total protein levels on the PVDF membrane after the transfer. Quantified results represent n = 3 samples. Statistical significance determined using Student’s t-test (P < 0.05).
Discussion
This study examined the hypoxic regulation of miRNA biogenesis proteins in cancer cells. Results showed that hypoxia down regulates the levels of Dicer and several other proteins (Drosha, TARBP2, DGCR8 and XPO5) in]volved in miRNA biogenesis. The hypoxic down regulation of Dicer protein levels was observed in multiple cell lines, suggesting that this could be a common phenomenon and not restricted to cancer cells. The reductions in Dicer protein levels were much more significant following longer durations of hypoxia (for 48 and 72 h) and with greater severity of hypoxic exposure (0.1% O
2). One likely explanation for this observation is high stability of Dicer protein. Therefore, the decrease in transcription might not be readily evident until 48 h. Consistent with this explanation, only modest reductions in Dicer protein levels were observed at shorter durations of hypoxic exposure. The mechanism of this hypoxic regulation appears to operate at several levels. The hypoxic suppression of Dicer mRNA levels appears to account for regulation in some (MCF7 and HT29) but not all cells (e.g. SKBR3) indicating the operation of other mechanisms in the hypoxic regulation of Dicer. During the course of this work, hypoxic repression of Dicer mRNA and protein was also reported by others [
34,
45,
47]. In exploring the mechanism of hypoxic regulation, we investigated the role of HIF and the HIF hydroxylases. In keeping with a role for the HIF hydroxylases in mediating this hypoxic repression, when cells were exposed to the HIF hydroxylase inhibitors, DMOG and desferrioxamine, a modest decrease in Dicer mRNA and protein levels was observed in MCF7 cells. However, somewhat surprisingly no, or very modest, effects on the repression of Dicer mRNA and protein levels were observed after HIF-1α and HIF-2α inhibition using siRNA treatments. This suggested that HIF did not play a major role in this hypoxic regulation of Dicer, a conclusion also reached by Ho et al. (2012) [
45], However, in a recent study Camps, et al. (2014) reported down regulation of Dicer mRNA levels in hypoxia in a HIF dependent manner [
48].
Our results indicate a mechanism of regulation that is oxygen and HIF hydroxylase dependent but independent of HIF. To explore this mechanism further, we examined for the possible contribution of each of the three HIF prolyl hydroxylases (PHD1, 2 and 3) and the HIF asparaginyl hydroxylase, FIH-1. When PHD2 was suppressed by RNA interference in normoxic conditions significant decreases in Dicer mRNA and protein expression were observed in normoxia, indicating a role for PHD2 in this hypoxic regulation. Interestingly, a recent study described a PHD2 dependent, but HIF independent, pathway for repressing protein synthesis in hypoxia. They reported of a PHD2 mediated phosphorylation of the eukaryotic elongation factor-2 and inhibition of protein translation under acute hypoxia [
49]. In keeping with such a mechanism, the other HIF hydroxylase enzymes (PHD1, PHD3) and FIH-1 were not involved in Dicer regulation by hypoxia.
In further exploring the mechanisms of hypoxic regulation and given the importance of miRNA mediated feedback loops in control of Dicer expression [
41,
42], we examined possible influence of hypoxically induced miRNAs in mediating hypoxic repression of Dicer expression. Hypoxically induced miR-210 is the best characterised example of a miRNA that shows substantial induction under hypoxic conditions [
32,
50] but had no significant influence on the levels of Dicer protein.
Two miRNAs, miR-103 and miR-107 have been shown to decrease miRNA biogenesis by targeting Dicer in cancer [
41] and are reportedly induced by hypoxia in some situations [
32]. To examine their potential contribution to hypoxic repression of Dicer, cells were exposed to hypoxia after inhibiting miR-103 and miR-107 with antagomirs. The repression of Dicer mRNA and protein levels in hypoxia was largely abrogated, consistent with a role for miR-103 and miR-107 in the hypoxic regulation of Dicer.
Other proteins involved in miRNA biogenesis (Drosha, TARBP2, DGCR8 and XPO5) also showed significant down regulation under hypoxia when compared to normoxia. A previous microarray study also showed modest but consistent hypoxic reductions in mRNA levels of genes (Dicer, TARBP2 and AGO2) involved in miRNA biogenesis [
36]. Ho
et al. (2012) also reported a decrease in DGCR8 and XPO5 protein levels in hypoxia, (though they did not see a decrease in Drosha levels in hypoxia) [
45]. This suggests the operation of a broader influence of hypoxia on the expression of genes encoding proteins that are essential in miRNA biogenesis. This is consistent with previous work indicating co-ordinate regulation of the levels of these proteins. When Dicer expression was suppressed there was a significant decrease in TARBP2 levels in keeping with previous reports [
43]. Similarly, when TARBP2 levels were reduced by siRNA treatment there was a modest decrease in Dicer protein. When Dicer and TARBP2 levels were examined over a time course of hypoxic exposure both proteins seemed to decrease co-ordinately. Previously others have shown a powerful influence of the levels of Dicer protein on the levels of TARBP2 and vice versa [
11,
43] and post transcriptional cross regulation between Drosha and DGCR8 [
51]. In this work we have also observed a further relationship linking Dicer and Drosha expression. Recent work has shown that Argonaute 2 stability is dependent on the availability of mature miRNAs. Dicer knockout influenced the mature miRNA production leading to decreased AGO2 stability [
52]. Similar mechanisms might be operating to co-ordinate the levels of other miRNA biogenesis proteins in hypoxia.
Even though there was a significant and consistent reduction in the levels of proteins with central roles in miRNA biogenesis machinery under hypoxia, we did not see a substantial effect of this on the expression levels of mature miRNAs over the time course of these experiments. Indeed under these conditions Dicer suppression by RNA interference was only associated with slight alterations in mature miRNA abundance. This was true both for miRNAs assessed by microarray studies and also when we focussed on examining the levels of particular miRNAs and their precursors by RT-PCR. Microarray analysis of MCF7 cells exposed to hypoxia showed eight miRNAs were significantly up regulated and four miRNAs were significantly down regulated. Of the eight up regulated miRNAs, miR-210 has been well validated as a highly induced miRNA in hypoxia [
35]. The lack of impairment of miR-210 maturation even after the decrease of many miRNA biogenesis proteins in hypoxia could be due to robustly increased transcription of the miR-210 gene in hypoxia. Recently deep sequencing of miRNA populations in hypoxic and normoxic HUVECs also showed no significant difference in the majority of miRNA species [
53]. Camps et al. (2014) reported down regulation of Dicer mRNA in hypoxia but failed to see a global down regulation of miRNAs in hypoxia compared to normoxia using small RNA sequencing techniques [
48]. They reported that 41 miRNAs were up regulated and 28 miRNAs were down regulated in hypoxia compared to normoxia in MCF7 cells [
48]. This work provides further support for Dicer down regulation by hypoxia but minimal influences on miRNA maturation.
In MCF7 cells only a modest decrease in mature and precursor levels of let-7a and miR-21 was seen after exposure to hypoxia, and we did not see an accumulation of pre-let-7a or pre-miR-21 in hypoxia. A significant reduction (P = 0.03) in mature miR-185 was observed in hypoxia in MCF7 cells, but there was no accumulation of precursors. These miRNAs (let-7a, miR-21 and miR-185) were chosen as previous reports showed they were Dicer dependent miRNAs and mature levels decreased with Dicer inhibition [
44‐
46].
The use of knockout murine models has provided compelling evidence for the essential role of Dicer in mature miRNA production. However, it has proven more difficult to clearly see such critical influences in a variety of experiments that have utilised siRNA mediated Dicer suppression [
44,
54]. Others have previously explored the surprising stability of some miRNAs with an average miRNA half-life of 119 h [
55], and this may largely explain the lack of changes in mature miRNA levels after 48 h of hypoxia or Dicer suppression by siRNA. The operation of Dicer independent pathways in the generation of miRNAs has been well described for a small number of specific miRNAs [
56‐
58] but seems a less plausible explanation for the lack of influence of Dicer manipulations on global alterations in miRNA abundance. Another possibility is that the levels of proteins such as Dicer are not rate limiting for the production of miRNAs under these conditions. These observations suggest that early and immediate responses to hypoxic stresses are not likely to be mediated by global alterations in mature miRNA levels.
During the course of this work, Ho et al. (2012) also described a reduction in Dicer levels after exposure to hypoxia, and reported a decrease in the levels of several mature miRNAs and an accumulation of their precursor-miRNAs in HUVEC cells [
45]. Whilst we were unable to replicate these findings using precisely controlled levels of hypoxia, this may suggest that in particular cells or conditions of hypoxic exposure, the hypoxic repression of Dicer can have important influences on miRNA biogenesis.
Whilst hypoxia exerted only modest effects on the production of mature miRNAs, we did see a significant influence of hypoxia on the operation of an exogenously introduced precursor miRNA. The processing and function of miR-200b was reduced in hypoxia when compared to normoxia. Dicer and TARBP2 proteins are involved in the processing of a precursor miRNA in to a mature miRNA and the reduction of both these proteins in hypoxia would likely have impacted on the maturation of the pre-miR-200b.
This work points towards the operation of non-HIF but HIF hydroxylase dependent pathways by which hypoxia controls reductions in gene expression. This effect of hypoxia appears to operate via several mechanisms of regulation and we have shown important influences of the HIF hydroxylase PHD2 and of the hypoxically regulated miRNAs miR-103 and miR-107 in this regulation. Furthermore, there appears to be co-ordinated regulation of several of the proteins involved in miRNA biogenesis under hypoxic and other conditions. However, influences on mature miRNA production were more modest than might have been anticipated from the substantial hypoxic reductions in protein expression of Dicer, Drosha, TARBP2 and XPO5. The breadth of this effect suggests a further and important interface between oxygen availability and gene expression.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
VB undertook all of the experimental work, prepared the results figures and analysis and drafted the manuscript. MM supervised the experimental work, participated in the design and interpretation of experiments and drafted the manuscript. JG conceived the study, its design and coordination, supervised the experimental work and drafted the manuscript. All authors read and approved the final manuscript.