Knock-Down of Argonaute 2 (AGO2) Induces Apoptosis in Myeloid Leukaemia Cells and Inhibits siRNA-Mediated Silencing of Transfected Oncogenes in HEK-293 Cells

Authors


Author for correspondence: Joon Myong Song, Research Institute of Pharmaceutical Sciences and College of Pharmacy, Seoul National University, Seoul 151-742, South Korea (fax +82 2 871 2238, e-mail jmsong@snu.ac.kr).

Abstract

Abstract:  Understanding the role of oncomirs allows new insights into the development of modern therapeutic approaches for the repression of multiple oncomirs in cancer cells. At present, no suitable approach is available to repress the development of multiple oncomirs in cancer cells. Herein, we report that argonaute 2 (AGO2) could be a unique molecule to regulate the development of multiple oncomirs in cancer cells. Knock-down of AGO2 by custom-made AGO2 siRNA resulted in the induction of apoptosis in myeloid leukaemia cells (HL-60). Further investigations revealed that knock-down of AGO2 by custom-made AGO2 siRNA in HEK-293 cells resulted in silencing of the expression of target genes vascular endothelial growth factor A and histone deacetylase 2, which are known to be involved in the development of myeloid leukaemia. From these results, it can be predicted that AGO2 could regulate siRNA-mediated RNAi pathways in cancer cells. Furthermore, we investigated the possible implication of AGO2 in drug-induced apoptosis. Investigations revealed that treatment with the newly synthesized drug analogue SH-03[{(7S,7aR,13aS)-9,10-dimethoxy-3,3-dimethyl-7,7a,13,13atetrahydro-3H-chromeno[3,4-b]pyrano[2,3-h]chromen-7-ol}] could induce AGO2-mediated apoptosis in myeloid leukaemia cells via intrinsic apoptotic pathways independent of Dicer.

MicroRNAs (miRNAs) are small, endogenous non-coding RNA molecules (approximately 22 bases) that can modulate the post-transcriptional regulation of many cellular genes [1]. Understanding the role of miRNAs in various pathological conditions has suggested new possible mechanisms in cancer development [2]. Half of the miRNA genes have been found at fragile chromosomal sites and in regions that are commonly deleted or amplified in a variety of human cancers [1,3]. This indicates an important role of miRNAs in the initiation and progression of human cancer. Such miRNAs are designated as oncogenic miRNAs (oncomirs) [4,5]. Biogenesis of oncomirs and their repression in malignant cells is an emerging subject of modern cancer research. Oncomirs are usually over-expressed in cancer cells and promote tumour development by inhibiting tumour suppressor genes involved in the regulation of the cell cycle [3]. Oncomir development is also reported to be associated with point mutations, genomic deletions/amplifications, random DNA methylation and genetic/epigenetic alterations in malignant cells [4]. miRNAs, encoded by the Let-7 family, was the first group of oncomirs reported to modulate the expression of oncogenes (including Ras). Recently, miR-17-92 oncomir was shown to be up-regulated in 65% of individuals with B-cell lymphoma [6,7]. Similarly, miR-125b-1, the homologue of Caenorhabditis elegans lin-4, was reported to be deleted in patients with leukaemia, breast, lung, ovarian and cervical cancers [8,9]. These studies support the role of these genes as oncomirs.

Because oncomirs are reported to be over-expressed in various cancers [3], repression of signature oncomirs during the initiation and progression of tumours could be a useful approach towards the development of novel cancer therapies. However, the basic difficulty in implementing this approach is repression of multiple oncomirs in the cancer cells. It is extremely difficult to repress multiple oncomirs in cancer cells as each oncomir may contribute independently to cancer development. Moreover, developing a specific inhibitor against each individual oncomir is a difficult and costly task. To date, no therapeutic approach has been established to repress multiple oncomirs in cancer cells. This may be because of the fact that the biogenesis of oncomirs is a complex process. In addition, there is the possibility of the existence of oncomirs that are not yet characterized but are involved in cancer development. Thus, it is important to explore those molecules that are involved in the RNAi pathway and can modulate the biogenesis of multiple oncomirs.

Argonaute 2 (AGO2) is a crucial element of the RNAi pathway and can direct both short interfering (siRNA) and miRNA-mediated gene regulation [10,11]. The AGO2-mediated RNAi pathway has already been reported [12] in which miRNAs transcribed as long primary transcripts are processed by nuclear RNase III Drosha, transported to the cytoplasm and processed by a Dicer-containing complex. Processed miRNAs are loaded onto an argonaute-containing complex, RNA-induced silencing complex (RISC), which can recognize the target mRNAs and inhibit protein expression. Of the eight argonaute members [13], only AGO2 can mediate specific endonucleolytic cleavage, leading to mRNA cleavage or translational repression of the target mRNA [14]. Recent investigations have revealed that AGO2 is required for the production of mature miRNA but not Dicer (another central element of the RNAi pathway) [14–16]. AGO2 has been reported to cleave siRNAs and pre-miRNAs [17]. Based on the above literature, AGO2 could be a unique molecule to modulate the development of many oncomirs and the siRNA-mediated RNAi pathway in cancer cells. Therefore, silencing AGO2 expression in cancer cells could be the key factor in modulating oncomir-mediated cancer development.

In the present investigation, the fate of AGO2 silencing in the siRNA-mediated RNAi pathway was monitored. Herein, for the first time, we show that knock-down of AGO2 can induce apoptosis in myeloid leukaemia cells. We also show that AGO2 knock-down by custom-made AGO2 siRNA in HEK-293 cells inhibits the siRNA-mediated silencing of two significant target genes, vascular endothelial growth factor A (VEGFA) and histone deacetylase 2 (HDAC2), involved in the development of myeloid leukaemia [18,19]. Apart from the role of AGO2 in the regulation of siRNA-mediated silencing, further investigations were carried out to elucidate its role in drug-induced apoptosis. Myeloid leukaemia cells were treated with a newly synthesized drug analogue (SH-03) at different concentrations. The extent of AGO2 down-regulation was found to be directly proportional to the increasing concentrations of SH-03 and the induction of apoptosis in HL-60 cells, independent of dicer cleavage. These results suggest that AGO2 could be a therapeutic target against myeloid leukaemia. AGO2 may be associated with repression of the biogenesis of multiple oncomirs in myeloid leukaemia cells.

Materials and Methods

Cell culture.  HL-60 and HEK-293 cell lines were procured from the Korean Cell Line Bank (KCLB®, Seoul, Korea) and cultured in RPMI-1640 (30041, J. R. Scientific, USA) and Dulbecco’s modified Eagle’s medium (DMEM, 11995; Gibco, Grand Island, NY, USA), respectively. Cultures were supplemented with heat-inactivated (10% v/v) foetal bovine serum (FBS, 16000-044; Gibco), 60 μg/ml penicillin (Sigma, St. Louis, MO, USA) and 100 μg/ml streptomycin (Sigma). The cells were cultured in 25-cm2 cell culture flasks (Nunclon; Rochester, NY, USA) and maintained at 37°C under 5% CO2 (US AutoFlow; NuAire, Plymouth, MN, USA).

DNA isolation and PCR and DNA sequencing.  Genomic DNA from the control human subject was isolated using a QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA, USA) as instructed by the manufacturer. Genomic DNA (200 ng) was used as a template to amplify target genes (HDAC2/VEGFA) with the respective specific primers, HDAC2F (50 pmol; 5′-CCGGAATTCCCCTGAATTTGACAGTCTCACC-3′) and HDAC2R (50 pmol; 5′-CGCGGATCCCACAATAAAACTTGCCCAGAAAAAC-3′) or VEGFAF (50 pmol; 5′-CCGGAATTCAAGGAAGAGGAGACTCTGCGCAGAGC-3′) and VEGFAR (50 pmol; 5′-GCGGATCCTAAATGTATGTATGTGGGTGGGTGTGTCTACAGG-3′), under the following temperature profile: 95°C for 4 min.; 30 cycles of 95°C for 30 sec., 50, 51.6, 53.2, 54.8 or 56.4°C for 30 sec. and 72°C for 1 min.; and a final extension at 72°C for 5 min. PCR amplification was confirmed by analysing the amplified fragments (173-bp HDAC2 and 208-bp VEGFA) on 1.2% agarose gel with standard molecular weight markers. Different annealing temperatures were used in a single gradient PCR thermal cycling profile so as to avoid unspecific amplification and to standardize the PCR profile in a single reaction.

Validations of amplified gene products were carried out using DNA sequencing. DNA sequence analysis of the VEGFA and HDAC2 genes was performed with the above-mentioned specific primers of the respective genes. Nucleotide blast analysis displayed a clear-cut homology of the amplified VEGFA/HDAC2 PCR products with the VEGFA/HDAC2 genes, respectively, in Homo sapiens. From these results, it can be assured that the amplified PCR products were HDAC2 and VEGFA genes.

DNA recombination and transformation.  PCR-amplified HDAC2 and VEGFA fragments were restricted with (EcoRI/NotI) enzymes. Restricted HDAC2 and VEGFA fragments were ligated in CoralHue® Monomeric Keima-Red (4687 bp; MBL, Woburn, MA, USA) and mammalian expression vector pTagYFP-C (4700 bp; Evrogen, Moscow, Russia), respectively, as suggested by the manufacturer. Ligated fusion gene vectors (YFP-VEGFA and RFP-HDAC2) were separately transformed into competent Escherichia coli DH5α strains through lipofectamine-based chemical transformation, as suggested by the manufacturer (Invitrogen, Carlsbad, CA, USA). Positive transformants were selected on LB agar plates supplemented with kanamycin (30 μg/ml). Plasmids from resistant colonies were isolated using a Plasmid Midi Kit (Qiagen).

siRNAs.  Pre-validated siRNA (Silencer® negative control) was procured from Ambion Inc., Austin, TX, USA. Custom-made siRNAs (siRNA VEGFA, siRNA HDAC2 and siRNA AGO2) were designed from the GeneAssist™ Pathway Atlas, Ambion. Gene sequences were selected from the NCBI database, and then, a blast search was conducted using siRNA target finder. GC content, position of the gene sequence, 3′ UU overhangs, loop sequence and stretches of T’s or A’s in the siRNA sequence were taken into consideration to achieve maximum silencing efficiency with the custom-made siRNAs. The sequences of custom-made VEGFA, HDAC2 and AGO2 siRNAs were AACTTCTGGGCTGTTCTCGCT, AAGGAGGCGGCAAAAAAAAAG and AAGGATATGCCTTCAAGCCTC, respectively. The specificities of custom-made siRNAs against their respective targets were confirmed by NCBI nucleotide blast analysis.

Co-transfections.  Cloned HDAC2 and VEGFA plasmids were co-transfected (in triplicate) into HEK-293 cells in the appropriate culture conditions. HEK-293 cells (5 × 105/well) were seeded in DMEM and FBS without antibiotics in six-well plate. On the next day, plasmids were co-transfected into HEK-293 cells (at 70% confluency) using lipofectamine-based chemical transfection (Lipofectamine 2000; Invitrogen) according to the manufacturer’s instructions. For knock-down studies, 20 pmol of siRNA (individual or in combination) was used during co-transfection. The samples were analysed using newly developed single-cell imaging cytometry.

SH-03 synthesis and treatment.  Synthesis of SH-03 has been described in detail in our earlier publication [20]. Before the drug treatment, HL-60 cells were grown overnight to achieve log-phase growth. Cells (4.5 × 105 cells/ml) were treated with different concentrations of SH-03 (10, 25 and 50 μM) for 6 hr. To begin with, we explored the possibility whether SH-03 can induce DNA fragmentation in HL-60 cells. DNA fragmentation in HL-60 cells was detected after 6 hr of SH-03 treatment. Hence, based on these results, a 6-hr time frame of treatment was chosen for apoptosis assay and western blot analysis.

Apoptosis assay through imaging cytometry.  After drug treatment, cells were washed and resuspended in calcium-enriched binding buffer (100 μl; BD Biosciences, San Jose, CA, USA) to allow annexin V binding to phosphatidylserine on the cellular membrane. The cell suspension was treated with annexin V-Biotin (5 μl; BD Biosciences) followed by incubation at room temperature for 15 min. in the dark. Unbound annexin V-Biotin was removed by washing. Cell pellet, resuspended in binding buffer (100 μl), was incubated with 10 nM Qdot-streptavidin conjugate (Cat No. Q10141MP; Invitrogen) and 10 μl of propidium iodide (PI; BD Biosciences) for 15 min. at room temperature in the dark. Quantum dots dissolved in binding buffer were vortexed before the staining procedure to avoid possible aggregation issues. The cell suspension was centrifuged and the pellet was resuspended in binding buffer (100 μl). Cells were further subjected to quantitative cellular imaging analysis.

Apoptosis assay through flow cytometry.  Flow cytometry was used to analyse SH-03-treated HL-60 cells. After the drug treatment, pellets containing approximately 1 × 105 cells were resuspended in 100 μl of calcium-enriched binding buffer. For staining, 5 μl of FITC-labelled annexin V (Apoptosis Detection Kit; BD Biosciences, Franklin Lakes, NJ, USA) and 5 μl of PI were added to the cell suspension and the cells were incubated at 4.0°C for 10 min. in the dark. Additional 400 μl of binding buffer was added at the end of the incubation and cells were subjected to flow cytometric analysis using a FACSCalibur instrument (BD Biosciences).

Quantitative high-content cellular imaging.  After the drug treatment, the cells were washed with phosphate-buffered saline (PBS), treated with accutase (Thermo Electron Corporation, Waltham, MA, USA) and incubated for 10 min. at 37°C. Data acquisition and quantitative analysis using high-content cellular imaging cytometry were carried out as described in detail in our earlier publication [21]. Briefly, Ar-ion laser (Melles Griot Laser Group, 35-LAP6431-220) beam of 488 nm was used for the excitation of fluorochrome and fluorescent proteins (YFP and RFP). An interference filter (Olympus America Inc., Lake Success, NY, USA, U-MGFPHQ) was used to purify incident light source. The Ar-ion beam was focused onto the cells and the fluorescence emissions from cells were collected, passed through an acousto-optic tunable filter (AOTF, Brimrose, TEAF10-0.45-0.7-S) and detected by a CCD camera (CoolSNAP cf mono, Phitometrics, A05F871008). AOTF is an electronically tunable spectral bandpass filter used for tuning the laser wavelengths over a finite laser wavelength range. AOTF was set up to scan the spectral region from 463 to 688 nm at each 3.75-nm interval with the scanning rate of 1 wavelength/sec. (maximum scanning rate of AOTF, 1 Hz). Fluorescence images were detected by the CCD camera as a function of AOTF sweeping frequency. For each sample, focused, slightly defocused, stack and background (without sample) images were captured at an identical xy axis position of the sample platform followed by background correction and threshold intensity distribution to a defocused image. Data analyses were carried out automatically using commercially available software (MetaMorph, Version 7.1.3.0; Molecular Devices Corp., Downingtown, PA, USA). Confocal micrographs of YFP-VEGFA- and RFP-HDAC2-transfected cells were chosen at 524 and 620 nm, respectively, based on the emission maxima of the YFP and RFP proteins.

Electrophoresis and western blotting.  Cells (1 × 106) were treated with 10, 25 and 50 μM SH-03 for 6 hr. After drug treatment, the cells were washed with PBS. Lysis-M solution (Roche, Manheim, Germany) was used to prepare whole-cell extracts. Electrophoresis was carried out by loading equal amounts of protein (60 μg) on 12% sodium dodecyl sulphate-polyacrylamide gel (SDS-PAGE). After electrophoresis, proteins were transferred to a polyvinylidene difluoride membrane (BioRad, Hercules, CA, USA). Appropriate primary antibodies [anti-AGO2; anti-cleaved Dicer; anti-poly [ADP(Adenosine diphosphate)-ribose] polymerase (PARP), anti-tubulin and anti-actin; Abcam, Cambridge, MA, USA] were used for western blot analysis. Membranes were blocked overnight at 4°C in PBS containing Tween 20 (0.05%) and bovine serum albumin (5%). Primary antibodies were detected using a 1:10,000 dilution of secondary peroxidase-conjugated antibody (Dako Ltd., High Wycombe, Bucks, UK). The signals were developed using an enhanced chemiluminescence detection kit (Roche, Basel, Switzerland).

Results

Knock-down of AGO2 can induce apoptosis in myeloid leukaemia cells.

To explore the possibility of AGO2-induced apoptosis, 20 pmol of custom-made AGO2 siRNA was transfected into HL-60 cells (fig. 1A,B) in triplicate. After 48 hr, cells were analysed using single-cell imaging cytometry [21]. Wavelength-based separation through cellular imaging cytometry clearly showed the induction of apoptosis after 48 hr in AGO2 siRNA-transfected cells (fig. 1A). Cellular images obtained at 525 and 620 nm indicate annexin V-Qdot-positive apoptotic cells and PI-stained necrotic cells, respectively. AGO2 siRNA-transfected cells showed 68.1% apoptosis, whereas control cells (cells not transfected with AGO2 siRNA; fig. 1C,D) did not show any apoptosis.

Figure 1.

 Induction of apoptosis in argonaute 2 (AGO2) knock-down HL-60 cells. Confocal micrographs of representative Ago2 knock-down (A and B) and control (C and D) cells were taken at 525 (apoptotic cells) and 620 nm (necrotic cells). Numerical values shown in the bottom of (A) and (B) represent per cent apoptotic and necrotic cells, respectively.

AGO2 knock-down inhibits siRNA-mediated silencing of target oncogenes.

Oncomirs are often over-expressed in cancer cells and may interfere with expression profiles of artificially transfected target genes. Therefore, a non-myeloid HEK-293 cell line was chosen as a model to gain better insights into the effects of AGO2 knock-down on the expression profiles of two important genes (VEGFA and HDAC2) involved in the development of myeloid leukaemia. Fusion gene vectors (YFP-VEGFA and RFP-HDAC2) were co-transfected in HEK-293 cells. Fig. 2 shows the confocal micrographs obtained via single-cell imaging cytometry. It can be seen that the custom-made VEGFA siRNA (fig. 2B) and HDAC2 siRNA (fig. 2C) individually knocked down the expression of VEGFA and HDAC2 proteins, respectively, compared with the control cells transfected with silencer® negative control (fig. 2A). Both the HDAC2 and VEGFA proteins were knocked down when a combination of HDAC2 and VEGFA siRNAs was co-transfected into the cells (fig. 2D). These results demonstrate that siRNA knock-down was specific to the corresponding target and led to the silencing of the respective gene. It was also clear that the silencing of one target gene did not interfere with the expression of the other target gene.

Figure 2.

 Multitarget analysis using single-cell imaging cytometry in HEK-293 cells. The fluorescence of YFP-VEGFA and RFP- histone deacetylase 2 (HDAC2) fusion proteins was detected at 524 and 620 nm, respectively. RFP: Red fluorescent protein; YFP: Yellow fluorescent protein. (A–D) entail representative YFP-VEGFA and RFP-HDAC2 co-transfected HEK-293 cells treated with control siRNA, VEGFA siRNA, HDAC2 siRNA and VEGFA siRNA + HDAC2 siRNA, respectively, at the time of co-transfection. VEGFA, vascular endothelial growth factor A.

Based on these initial findings, further investigations were conducted to elucidate the effect of AGO2 siRNA on the expression profiles of VEGFA and HDAC2 proteins. Confocal micrographs in fig. 3 show that the expression of VEGFA and HDAC2 proteins remained unchanged after transfection of either AGO2 siRNA (fig. 3A) or the combination of AGO2, HDAC2 and VEGFA siRNAs (fig. 3B) in HEK-293 cells. The absence of AGO2 siRNA resulted in the knock-down of both the VEGFA and HDAC2 proteins by their respective VEGFA and HDAC2 siRNAs (fig. 3C). These results show that siRNA-mediated silencing of target genes does not work in the absence of AGO2 and that AGO2 knock-down alone does not interfere with the vector-mediated expression of fusion genes.

Figure 3.

 Effect of argonaute 2 (AGO2) siRNA on the silencing of vascular endothelial growth factor A (VEGFA) and histone deacetylase 2 (HDAC2) genes in HEK-293 cells. (A–C) entail representative YFP-VEGFA and RFP-HDAC2 co-transfected HEK-293 cells treated with AGO2 siRNA, AGO2 siRNA + VEGFA siRNA + HDAC2 siRNA and VEGFA siRNA + HDAC2 siRNA, respectively, at the time of co-transfection.

SH-03-induced apoptosis could be mediated through AGO2 in leukaemia cells.

Apart from the AGO2 knock-down studies, the fate of AGO2 in drug (SH-03)-treated cells was also investigated. The apoptosis-inducing potential of the SH-03 drug analogue via PI3K/AKT pathway in HL-60 cells has already been reported by our group [21]. AKT is a serine/threonine protein kinase that has a key role in various cellular processes. HL-60 cells were treated with different concentrations (10, 25 and 50 μM) of SH-03 for 6 hr. After the drug treatment, half of the cells were used to study the expression of AGO2 and half of the cells were used to monitor the induction of apoptosis. Western blot analysis (fig. 4A) showed that the extent of AGO2 down-regulation was sequentially proportional to the increasing concentrations of SH-03 compared with that in control cells. The expression of actin (positive control) remained the same in control and SH-03-treated cells. Flow cytometric analysis revealed that the rate of apoptosis was increased with increasing concentrations of SH-03. HL-60 cells treated with 50 (fig. 4B), 25 (fig. 4C) and 10 μM (fig. 4D) SH-03 showed 30.8%, 25.1% and 19.5% apoptosis, respectively. These results indicate that the extent of AGO2 down-regulation was directly proportional to the induction of apoptosis in HL-60 cells.

Figure 4.

 (A) Western blot analysis of argonaute 2 protein in HL-60 cells after treatment with 10, 25 and 50 μM SH-03 (6 hr). (B–D) illustrate representative dot plots obtained from flow cytometric analysis and show the induction of apoptosis in HL-60 cells after the treatment of 10, 25 and 50 μM SH-03, respectively.

SH-03-triggered apoptosis could be AGO2-mediated and Dicer-independent.

Further investigation was conducted to determine the involvement of another core component (Dicer) of the RNAi pathway in the Ago-2-mediated apoptosis pathway. HL-60 cells were treated with 10, 25 and 50 μM SH-03 for 6 hr. Western blot analysis (fig. 5) showed that the expression of poly (ADP-ribose) polymerase (PARP) was with increasing concentrations of SH-03. Expression of Dicer remained the same in all samples (control and SH-03-treated cells). The expression of tubulin (positive control) remained the same in control and SH-03-treated cells. This implies that Dicer was not cleaved after SH-03 treatment. Thus, it can be concluded SH-03-triggered apoptosis HL-60 cells could be mediated through AGO2-mediated pathway which is independent of Dicer cleavage.

Figure 5.

 Western blot analysis of PARP and Dicer proteins in HL-60 cells after treatment with 10, 25 and 50 μM SH-03 (6 hr).

Discussion

miRNAs are an abundant class of gene regulators that have been reported to regulate a wide range of biological functions such as cellular differentiation, proliferation and apoptosis [4]. High-throughput studies with tumour samples have revealed that oncomir profiling can distinguish between different cancer signatures and anticipated patient outcomes with considerable accuracy [4,22]. Oncomirs are over-expressed in many cancer types [3]. Therefore, efforts have been made to repress the biogenesis of signature oncomirs in cancer cells. The basic bottleneck in implementing this approach is the existence of multiple oncomirs (both known and unknown) in the cancer cell that contribute to cancer development [22]. Thus, this approach can only be implemented if all the oncomirs can be repressed in the cancer cells. This might require high-throughput target analysis combining genomics, miRomics and proteomics to delineate the spectrum of targets [4]; however, such an approach would be both costly and time-consuming.

At this time, no method is available to repress the development of multiple oncomirs in malignant cells. In the present study, we chose AGO2 as a potential target molecule and studied the ability of AGO2 to modulate the siRNA-mediated RNAi pathway. Our results provide initial evidence that the knock-down of Ago2 can induce apoptosis in myeloid leukaemia cells (fig. 1). Available literature illustrates that AGO2 is a core component of RISC [23], which can regulate the alterations in gene expression by siRNA [24] and miRNA [16] mediated RNAi pathway. Specific catalytic cleavage by AGO2 is required for the production of mature miRNA [16] and target recognition by siRNA. Thus, to regulate the intracellular gene expression, siRNAs or miRNAs have to undergo catalytic cleavage by the AGO2 molecule. These facts led us to predict that AGO2 could be a unique molecule to regulate oncomir development as well as the siRNA-mediated RNAi pathways in cancer cells. Argonaute proteins are reported to be over-expressed in cancer cells [23]. This is an additional advantage in selecting AGO2 as a therapeutic target because AGO2 inhibitors specifically targeting the intracellular AGO2 in cancer cells could be developed.

Based on our prediction, further investigations were carried out to elucidate the effect of AGO2 on the siRNA-mediated silencing of the target genes. HDAC2 and VEGFA genes were selected as the target genes based on their potential role in the development of leukaemia. LBH589, an HDAC inhibitor, has been shown to possess antileukaemia activity via depletion of EZH2 and DNA methyltransferase 1 through the disruption of chaperone association with heat shock protein (hsp) [18]. Similarly, HDAC inhibitors (MS-275 and LAQ824) had antileukaemia activities [25,26]. Recently, expression of VEGF and its receptors was found on chronic lymphocytic leukaemia B cells, which were reported to interact with signal transducers and activators of transcription (STAT) 1 and 3 genes [27]. VEGF was also reported to induce leukaemia cell migration via P38/ERK1/2 kinase pathway [19]. VEGF (165) was reported to promote survival of leukaemic cells by Hsp90-mediated induction of Bcl-2 expression and apoptosis inhibition [28]. Our results showed that AGO2 knock-down in HEK-293 cells inhibits siRNA-mediated silencing of HDAC2 and VEGFA genes (fig. 3). These results represent an important step towards understanding the role of AGO2 in the siRNA-mediated silencing pathway.

Fig. 3A shows that HDAC2 and VEGFA expression remained unchanged in HEK-293 cells after AGO2 knock-down. Both the mammalian expression vectors used in this study contain a CMV promoter. This promoter provides strong, constitutive expression of TagYFP or TagRFP fusion genes in the cell. The SV40, PUC and f1 origins of these vectors facilitate the expression of the SV40 T-antigen, replication in E. coli and single-stranded DNA production in the host cells, respectively. Thus, fusion genes can be constitutively expressed in the host cells despite interference rendered by any other intracellular mechanism. Because AGO2 knock-down alone did not interfere with the vector promoter-mediated expression of the fusion genes (fig. 3A), we may infer that AGO2 specifically interferes with the RNAi-mediated pathway.

miRNAs could function as tumour suppressors as well as oncogenes [1]. Therefore, one possible question is the fate of tumour suppressor miRNAs in cancer cells after AGO2 knock-down. It has been reported that cancer cells have elevated levels of oncogenes that regulate the genes involved in cell cycle regulation [29]. Most of the tumour suppressor genes (e.g. p53), as well as tumour suppressor miRNAs involved in cell cycle regulation, are mutated in cancer cells [30]. miR-15a and miR-16-1 miRNAs are reported to negatively regulate an antiapoptotic protein (BCL2) expression in cancer cells, including leukaemias and lymphomas [31]. Deletion or down-regulation of miR-15a and miR-16-1 is thought to be related with leukaemogenesis and lymphomagenesis [31]. Similarly, mature miRNA levels of miR-143 and miR-145 were reported to be reduced in colorectal tumours [32]. Therefore, efforts to repress the expression of oncogenes in cancer cells is a much easier approach than efforts to repair the cell cycle machinery by activating tumour suppressor genes in cancer cells. Moreover, there are other cellular mechanisms apart from the miRNA-mediated tumour suppression that can activate tumour suppressor genes in cancer cells.

Our investigation also illustrates the role of AGO2 in drug-induced apoptosis. Fig. 4 describes the effect of a drug analogue (SH-03) on AGO2 in myeloid leukaemia cells. SH-03 treatment resulted in dose-dependent induction of apoptosis that was directly proportional to the extent of AGO2 down-regulation in HL-60 cells. The detailed information about the postulated mechanism of action of SH-03 in cancer cells has been mentioned in detail in our earlier publication [33]. Briefly, SH-03 is a rotenoid-containing deguelin analogue which is reported to activate Hsp90, intracellular hypoxia-inducible factor 1 subunit a, STAT proteins and mammalian target of rapamycin in non-small cell lung cancer (NSCLC) and malignant human bronchial epithelium cell lines [20]. In myeloid leukaemia cells, SH-03 triggers death signals via the caspase-9-mediated intrinsic apoptotic pathway by inhibiting the PI3K/AKT signal transduction pathway [33]. The results obtained from the present investigation entail that AGO2 expression was sequentially down-regulated by increasing concentrations of SH-03 compared with that in control cells. PI3K/AKT signal transduction and AGO2-mediated silencing are two independent events and their correlation, if any, is currently unclear.

Fig. 5 shows the postulated mechanism of SH-03-triggered apoptosis in HL-60 cells via an AGO2-mediated, Dicer-independent pathway. Caspase-mediated Dicer cleavage has been reported in the TNFα-mediated apoptosis in HeLa cells [34]. This suggests that Dicer cleavage occurs after the initiation of apoptotic signals in HeLa cells. In our findings, AGO2 initiated apoptosis in HL-60 cells. Thus, AGO2 was involved in the initiation of apoptotic signals in HL-60 cells. It would be interesting to correlate these results with the recent findings that AGO2 and not Dicer is involved in miRNA production (miRNA) [14–16]. AGO2, being a unique molecule, could repress development of multiple oncomirs in cancer cells, thus triggering the initiation of apoptotic signals. For better statistical understanding, three independent experiments were performed for elucidating the induction of apoptosis in AGO2 knock-down HL-60 cells, multitarget analysis using single-cell imaging cytometry in HEK-293 cells, understanding the effect of AGO2 siRNA on the silencing of VEGFA and HDAC2 genes in HEK-293 cells and western blot analysis of AGO2, PARP and Dicer proteins in HL-60 cells.

Available literature suggests that siRNAs can function as miRNAs [35]. In the present investigation, we therefore designed an experimental model based on the AGO2-mediated silencing of the siRNA pathway pertaining to two target oncogenes to predict the fate of oncomir development in AGO2 knock-down cancer cells. Fig. 6 shows the schematics of our approach and the possible mechanism of the induction of apoptosis owing to the repression of oncomirs in AGO2 knock-down cancer cells. In the oncomir development pathway, pre-miRNA exported from the nucleus is processed by AGO2 and RISC to form mature oncogenic miRNA that binds to the target mRNA in the cancer cell. Conversely, AGO2 knock-down cells could not process pre-miRNAs exported from the nucleus, resulting in the repression of oncomirs. Repression of multiple oncomirs in the cancer cell leads to the induction of apoptosis. In the siRNA-mediated silencing pathway, transfected siRNAs are processed by AGO2 and RISC to form cleaved siRNA, which binds to the target mRNA in the cancer cell. AGO2 knock-down inhibits siRNA cleavage; as a result, the siRNA is degraded and unable to bind to the target mRNA.

Figure 6.

 Schematics of the possible mechanism involved in the repression of oncomirs/siRNAs via argonaute 2 knock-down in cancer cells. Two dotted circles represent the mechanism involved in the oncomir development and siRNA-mediated silencing.

In conclusion, we suggest that AGO2 could be a possible therapeutic target against myeloid leukaemia. This may represent a new opportunity for repressing oncomir development. The above-postulated mechanisms are at the primitive stage and need to be extensively validated with the help of in vitro and in vivo models. More thorough investigations could also imply similar roles for AGO2 in other types of cancer. Our future work will correlate the findings obtained in HEK-293 system with in vitro (other myeloid leukaemic cell lines) or in vivo systems to find out whether there is any correlation between various cellular pathways targeted by SH-03. Further insight into the role of other proteins involved in the RNAi pathway will be important while designing Ago2-specific inhibitors to target cancer cells.

Acknowledgement

This work was supported by the National Research Foundation of Korea (NRF) Grant funded by the Ministry of Education, Science and Technology (MEST) (2010-0017903, R11-2008-044-01004-0) and a grant of the Korea Healthcare technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea (A100096).

Conflict of Interest Declaration

The authors report no conflict of interests. The authors alone are responsible for the content and writing of the paper.

Ancillary