Independent combinatorial effect of antisense oligonucleotides and RNAi-mediated specific inhibition of the recombinant Rat P2X₃ receptor

Abstract

Synthetic 21-bp-long short interfering RNAs (siRNA) can stimulate sequence-specific mRNA degradation in mammalian cell cultures, a process referred to as RNA interference (RNAi). In the present study, the potential of RNAi was compared to the traditional antisense approach, acting mainly via RnaseH, for targeting the recombinant rat pain-related cation-channel P2X₃ expressed in CHO-K1 and a rat brain tumour-derived cell line, 33B. Downregulation of the P2X₃ receptor was evaluated at the mRNA, protein, and functional levels. In this study, four siRNA duplexes induced up to 95% sequence-specific inhibition of the P2X₃ mRNA, independent of the type of 2 nt 3′-overhang modification and the location of the targeted sequences. Furthermore, we detected and characterised an independent combinatorial effect of antisense oligonucleotides (ASOs) and RNAi-mediated specific inhibition of the P2X₃ receptor. Enhanced downregulation was observed only when siRNA was combined with nonhomologous ASO, targeting distant regions on the common P2X₃ mRNA. The two reagents resulted in more efficient downregulation of P2X₃ mRNA when administered in combination rather than separately. To our knowledge, this is the first investigation at the molecular level of the potential benefits of mixed antisense and RNAi-mediated treatment for inhibiting expression of a medically relevant pain-related gene.

INTRODUCTION

Extracellular adenosine triphosphate (ATP) is implicated in multiple sensory processes. Early experiments demonstrated that ATP is released from sensory nerves (1) and produces fast excitatory potentials in dorsal root ganglia (DRG) neurons (2). More recent work has shown that extracellular ATP signalling involves two families of proteins: the metabotropic G protein-coupled P2Y receptors and the P2X receptors, which function as ligand-gated cation channels (reviewed in 3,4). The P2X receptor family includes seven closely related members, widely expressed throughout both the central (CNS) and peripheral nervous systems (PNS) and in peripheral tissue (5). Among the P2X receptors, P2X₃ attracts special attention due to its role in pain signalling (reviewed in 6), which has been established in numerous studies. In contrast to the other members, the P2X₃ receptor is highly localised to peripheral sensory neurons in DRG (7), where a functional receptor forms by homo- or heterodimerisation with P2X₂ (8). P2X₃ expression is upregulated following chronic constriction injury of the sciatic nerve (9), which also provokes an ectopic sensitivity to ATP (10). P2X₃ knockouts show lack of rapidly desensitising currents induced by ATP and significant reduction in pain reception in response to ATP (11,12). Unfortunately, there are no highly specific compounds available for differentiating between P2X receptors, which complicates functional studies of this receptor family (13).

The absence of a specific inhibitor for the P2X₃ receptor led us to develop an alternative approach based on antisense technology. Since the early idea (14) and the first elegant demonstration (15) of antisense oligonucleotides (ASOs) as a method for the selective inhibition of gene expression, various parameters have been introduced into ASO chemistry to improve their performance. For example, oligonucleotide chimeras comprising blocks of phosphorothioate DNA or alkylated ribose derivatives (16), including 2′-O-(2-methoxyethyl)-ribonucleoside (MOE) (17), are now considered the best reagents for many applications. Meeting the requirements of enhanced stability, increased affinity for target mRNA, and the recognition and induction of RNaseH activity, chimeric ASOs have been used successfully to downregulate the expression of numerous targets in the CNS (18). The effectiveness of complementary oligonucleotides is usually restrained by natural folding of the targeted mRNA and diverse screening procedures are in use for the selection of a potent antisense inhibitor. Recently, Dorn et al. (19) observed potent sequence-specific downregulation of P2X₃ expression with two (ASO-5037 and 5038) out of 11 tested MOE-modified phosphorothioate chimeras against the rat P2X₃ receptor expressed in CHO-K1 cells. In another study, Barclay et al. (20) evaluated the effect of the selected chimeric P2X₃ ASO, during continuous intrathecal administration in the rat. Using animal model systems, the authors demonstrated that selective antisense-based blockage of P2X₃ expression leads indeed to relief from chronic neuropathic and inflammatory pain. However, in view of the well documented non-specific toxicity at high doses and reversibility of the therapeutic effects at low doses (21), the prospects for ASO in the effective relief of both inflammatory and neuropathic pain require careful evaluation.

As an alternative to the antisense approach, gene expression can be specifically inhibited by homologous double-stranded RNA. The process, referred to as RNA interference (RNAi), was originally discovered and has been most extensively studied in Caenorhabditis elegans (22–25) and Drosophila melanogaster (26–32). Biochemical studies in Drosophila cells and embryo extracts led to the discovery of mechanistic details of RNAi, showing that long dsRNA is processed by a nuclease named ‘dicer’ into 21–23 nt duplexes, termed short interfering RNA (siRNA). Subsequently, siRNAs associate with an RISC nuclease complex and guide this multicomponent enzyme to degrade mRNA in a sequence-specific manner. A similar process can occur in mammalian early embryos and embryonal cell lines (33–38). In cultured mammalian cell lines, however, uninterrupted RNA duplexes longer that 30 bp trigger unspecific cellular responses through the activation of dsRNA-dependent protein kinase PKR and RnaseL, effectors of interferon-induced cell death (39). Stimulation of the IFN pathway can be avoided in many mammalian cell lines by direct administration of 21 nt siRNAs, which then elicit only sequence-specific RNAi-mediated inhibition of gene expression (40,41). Synthetic siRNA duplexes introduced by cationic lipid-mediated transfection, electroporation or microinjection as well as intracellular expression of siRNAs from plasmid DNA (42–46, reviewed in 47) are now commonly used laboratory methods. The natural biological process of RNAi thus provides a new approach for downregulation of gene expression in mammalian systems.

In this study, four synthetic siRNA duplexes were characterised for their ability to inhibit specifically the recombinant rat P2X₃ receptor expressed in a Chinese hamster ovary cell line (CHO-K1) and in the rat brain tumour-derived cell line 33B. The potential of RNAi was compared to the traditional antisense approach and results were analysed on the mRNA, protein and functional levels. To evaluate whether co-treatment with the two reagents results in improved performance, combinations of ASO and siRNAs were used to downregulate P2X₃, an established target in different forms of pain. We believe that these experiments are the first to demonstrate, at the molecular level, the potential benefits of combined antisense- and RNAi-mediated treatment.

MATERIALS AND METHODS

Cell lines

The rat brain tumour cell line 33B expressing recombinant rat P2X₃ (X91167) was prepared exactly as the CHO-rP2X₃ cell line, described previously (19). The Chinese hamster ovary cell line CHO-K1 expressing both recombinant rat P2X₂ and P2X₃ proteins was prepared by double transfection with respective cDNA clones. P2X₃ cDNA was amplified on total DRG RNA by RT–PCR (six cycles of 30 s at 94°C; 60 s at 62°C, 180 s at 68°C, followed by 29 cycles of 30 s at 94°C, 240 s at 68°C) using Advantage-HF2 Polymerase (Clontech) and the following forward and reverse primers, respectively: 5′-cgcaagcttggctgtgagcagtttctcagtatgaacttg-3′ and 5′-cttgagctcgggaagaggccctagtgaccaatag-3′. The PCR product was initially cloned in pGEM T-Easy (Promega), confirmed by sequencing, and then subcloned as the NotI fragment into pcDNA5/FRT. The P2X₂ sequence (U14414) was subcloned into the pcDNA5/FRT-Neo as a cDNA flanked by BamHI and XhoI restriction sites (gift of B. Fakler, University of Tubingen). The CHO-rP2X₂/P2X₃ cell line was created by a two-step transfection: CHO Flp-In cells with two integrated FRT sites (Invitrogen) were transfected with pcDNA5/FRT-Neo-P2X₂ using FuGENE 6 reagent (Roche) and selection with 500 µg/ml Geneticin, followed by transfection with pcDNA5/FRT-P2X₃ and selection with 200 µg/ml Hygromycin-B.

Antisense oligonucleotides and RNA duplexes

The ASO against rat P2X₃ (accession no. X91167) described previously (19) and the MOE-modified RNA oligonucleotides were synthesised on ABI394 or Expedite/Moss Synthesizers (Applied Biosystems) using phosphoramidite chemistry (48), HPLC purified, and characterised by electrospray mass spectrometry and capillary gel electrophoresis (data not shown). Deoxythymidine-modified RNA oligonucleotides (RNA-4882 to -4885) were purchased from Xeragon. Annealing of siRNAs was performed in hybridisation buffer (30 mM HEPES/KOH, pH 7.4; 100 mM K-acetate; 2 mM Mg-acetate), exactly as described by Elbashir et al. (40).

Cationic lipid and liposome-mediated transfections

Cells were transfected on 24 well plates using Oligofectamine or Lipofectin reagent [Invitrogen (earlier Gibco BRL)]. For transfections with Oligofectamine, cells (2–5 × 10⁴ per well) were seeded in 500 µl of DMEM supplemented with 10% FBS without antibiotics. The next day, the culture medium was replaced with 50, 200 or 500 µl of fresh OptiMEM (containing neither serum nor antibiotics) and subsequently overlayered with 100 µl of the freshly prepared standard Oligofectamine transfection mixture (41), including 60 pmol of siRNA or ASO. The corresponding final concentrations of siRNA or ASO reagents were 400, 200 and 100 nM. After initial incubation for 4 h at 37°C, samples were adjusted if necessary with Optimem to 600 µl and supplemented with 70 µl of FBS (10% final concentration). Following subsequent incubation at 37°C in a humidified incubator with 5% CO₂ for 24 or 48 h, 70–90% confluent samples were collected after 24 h for RNA and after 48 h for protein extraction or functional analysis (otherwise as indicated in figures). Lipofectin-mediated transfections were performed exactly as described by Dorn et al. (19).

Electroporation of mammalian cells

Cells were transfected with 0.15, 0.3, 0.6 or 1.2 nmol of ASO or siRNA duplex using standard electrotransfection (10⁶ cells/125 µl in a BioRad gene pulser cuvette 0.4 cm, 250 V, 0.3 µF, infinite resistance). Following electroporation, samples were immediately combined with 6 ml of the culture medium. The corresponding final concentrations of ASO or siRNA reagents were 10, 50, 100 and 200 nM. Cells were plated on uncoated 24 or 96 well plates (Costar) and incubated at 37°C for 24 h or 48 h, followed by RNA or protein extraction, respectively, or used for functional analysis.

Total RNA isolation and assay by quantitative real-time PCR (Q-PCR)

Total RNA was extracted and purified using RNeasy 96 kit (Qiagen). Primer pairs and FAM/TAMRA-labelled TaqMan probes for Q-PCR were designed using the Primer Express v 2.0 program (ABI PRISM, PE Biosystems). For the Q-PCR, 50 ng of total RNA was mixed with 5′ and 3′ primers (Table 1; T-forward and T-reverse; 10 µM each), Taqman probe (Table 1; Taqman; 5 µM), MuLV reverse transcriptase (6.25 U; PE Biosystems), RNase Out RNase inhibitor (10 U; Invitrogen) and the components of the TaqMan PCR reagent kit (Eurogentec) in a total volume of 25 µl following the TaqMan PCR reagent kit protocol (Eurogentec). Reverse transcription and Q-PCR were performed in a GeneAmp Sequence Detector 5700 (PE Biosystems) as follows: 2 min reverse transcription at 50°C, 10 min denaturation at 95°C followed by 50 cycles of denaturation for 15 s at 95°C and annealing and elongation for 1 min at 60°C. The relative gene expression was calculated as described in the ABI PRISM 7700 user bulletin no. 2 (PE Biosystems). Data in the figures represent average values from a minimum of three experiments. All samples were normalised using Pre-Developed TaqMan Assay Reagents for detection of the 18S rRNA level (Applied Biosystems) treated as an internal control.

Table 1. Sequences of Rat P2X₃ DNA and RNA oligonucleotides.

Number	Position	Sequence	Modification	Comment
ASO-10326	313–330	5′-GTCCCTCACttggtaggc-3′	All phosphorothioate/n = 2′-MOE	Antisense
ASO-5037	956–973	5′-CTCCATCCAgccgagtga-3′	All phosphorothioate/n = 2′-MOE	Antisense
ASO-MM-5655	956–973	5′-CTACAGCCAtccgcgtga-3′	All phosphorothioate/n = 2′-MOE	4 MM
ASO-5038	1033–1050	5′-CTCGCTGCCgttctccat-3′	All phosphorothioate/n = 2′-MOE	Antisense
ASO-5596	Synthetic	5′-ccttaCCTGCTAGctggc-3′	n = phosphorothioate/n = 2′-MOE	Scrambled
RNA-7556	956–976	5′-UCACUCGGCUGGAUGGAGUaa-3′	n = phosphorothioate/2′-MOE	Sense
RNA-7557	954–974	5′-ACUCCAUCCAGCCGAGUGAaa-3′	n = phosphorothioate/2′-MOE	Antisense
RNA-MM-7558	956–976	5′-UCACUGCGCUCGAUGCAGUaa-3′	n = phosphorothioate/2′-MOE	4 MM
RNA-MM-7559	954–974	5′-ACUGCAUCGAGCGCAGUGAaa-3′	n = phosphorothioate/2′-MOE	4 MM
RNA-8646	956–976	5′-UCACUCGGCUGGAUGGAGUuu-3′	n = phosphorothioate/2′-MOE	Sense
RNA-8647	954–974	5′-ACUCCAUCCAGCCGAGUGAag-3′	n = phosphorothioate/2′-MOE	Antisense
RNA-4882	956–976	5′-UCACUCGGCUGGAUGGAGUtt-3′	n = phosphorothioate/2′-deoxy-T	Sense
RNA-4883	954–974	5′-ACUCCAUCCAGCCGAGUGAtt-3′	n = phosphorothioate/2′-deoxy-T	Antisense
RNA-4884	312–332	5′-GGCCUACCAAGUGAGGGACtt-3′	n = phosphorothioate/2′-deoxy-T	Sense
RNA-4885	310–330	5′-GUCCCUCACUUGGUAGGCCtt-3′	n = phosphorothioate/2′-deoxy-T	Antisense
RNA-7126	Synthetic	5′-ACGGCAGCGUGCAGCUCGCCga-3′	n = phosphorothioate/2′-MOE	Unrelated
RNA-7127	Synthetic	5′-GGCGAGCUGCACGCUGCCGUcc-3′	n = phosphorothioate/2′-MOE	Unrelated
T-forward	1034–1052	5′-TGGAGAACGGCAGCGAGTA-3′	None	Sense
Taqman	1054–1077	5′-CGCACACTCCTGAAGGCTTTTGGC-3′	5′-FAM/3′-TAMRA	Sense
T-reverse	1079–1097	5′-ACCAGCACATCAAAGCGGA-3′	None	Antisense

Underline marks indicate mismatches. Capital and small letters represent unmodified and modified nucleotides, respectively.

Western blotting

Cells grown in 24 well plates were washed with PBS and lysed as described previously (19). A 10 µg aliquot of solubilised proteins was separated in SDS–PAGE using NuPAGE™ 4–12% Bis-Tris Gels in a NOVEX™ Xcell SureLock Mini-Cell system, followed by transfer to Invitrolon™ PVDF membrane (Invitrogen). Blots were reversibly stained with 1% solution of Ponceau S (Pierce) to check for the equal loading and transfer. Blocking and immunodetection was performed with the ECL Western Blotting Kit (Amersham Pharmacia Biotech) using rabbit anti-rat P2X₃ antibody (Neuromics). This antibody detects multiple glycosylated forms of the P2X₃ protein, migrating at ∼50 and 75 kDa. Blots were scanned and quantified with ImageQuant™ (Molecular Dynamics). Experiments were performed twice yielding similar results.

Ca²⁺-measurements with FLIPR (fluorescence immuno-plate reader)

For FLIPR experiments, cells were loaded with fluo-4 AM (Molecular Probes) in the presence of 2.5 mM probenicid for 30–45 min, washed twice with Hank’s balanced salt solution (HBSS; Invitrogen) and 20 mM HEPES, and transferred to the fluorescence reader (FLIPR, Molecular Devices). Drug plates were prepared at five times the final concentration. Fluo-4 fluorescence was measured at the rate of 0.5 Hz for 3 min. Agonist α,β-methylene-ATP was applied after 20 points baseline detection.

FLIPR sequence files were analysed using Igor Pro (Wavemetrics). The baseline was set as the average of 20 points before α,β-methylene-ATP addition, and the peak was detected as the maximal signal in the 50 data points after α,β-methylene-ATP addition. Relative change of fluorescence (dF / F) was determined as (peak – baseline) / (baseline) values. Data are presented as means ± SEM.

RESULTS

Antisense reagents, including 18mer DNA ASO and 21mer siRNA duplexes were designed to target the coding sequence of the recombinant rat P2X₃ gene (Table 1 and Fig. 1). ASO-5037 and ASO-5038 were characterised previously as potent inhibitors of P2X₃ expression upon transfection with Lipofectin (19). In the current study, the same sequences were selected to target the P2X₃ gene with RNAi. Two siRNA duplexes, siRNA-7556/7557, and its four mismatch analogue, siRNA-MM-7558/7559, were compared to the corresponding ASO-5037 and ASO-MM-5655 in two cell lines (CHO-rP2X₃ and 33B-rP2X₃) expressing recombinant targeted sequence. In both cell lines, Oligofectamine-delivered siRNA-7556/7557 affected P2X₃ mRNA downregulation in a dose-dependent manner (Fig. 2A). Twenty-four hours after transfection, 200–400 nM siRNA-7556/7557 resulted in 60–70% inhibition of P2X₃ mRNA, compared with the mismatch and mock controls (siRNA-MM and Oligofectamine alone, respectively). Tested in parallel, neither siRNA delivered with Lipofectin nor ASO delivered by Oligofectamine (tested only in CHO-rP2X₃) were effective. Downregulation by siRNA-7556/7557 delivered by transfection with Oligofectamine was confirmed at the protein level (Fig. 2B) by western blotting and immunodetection using an anti-P2X₃ antibody. At 72 h after transfection in our system, 100 nM siRNAi-7556/7557 induced 75% downregulation of the recombinant P2X₃ protein.

Figure 1.

(A) Schematic representation of the P2X₃ receptor mRNA with the ASO and siRNA reagents selected to downregulate P2X₃ gene expression. P2X₃ mRNA is represented by a black continuous line with the ends of the P2X₃ coding region indicated by boxes. siRNA duplexes are shown by double contrapositioned arrows located at the top, each representing sense (right headed) and antisense (left headed) single strands. ASOs are indicated by single left-headed arrows at the bottom, and crosses represent complementarity between siRNA and overlapping homologous ASO. (B) Sequences and base-pairing of the siRNA duplexes used in the present study. Underline marks indicate mismatches.

Figure 2.

RNAi-mediated inhibition of P2X₃ expression. (A) Dose- dependent mRNA inhibition by siRNA-7556/7557 versus its mismatch analogue siRNA-MM-7558/7559. Oligofectamine- and Lipofectin-mediated transfections are separated by a space and grouped left and right, respectively. Concentrations of ASO and siRNAs as indicated. Twenty-four hours after transfection of CHO-rP2X₃ (white bars) and 33B-rP2X₃ cells (black bars), P₂X₃-specific mRNA was measured with Q-PCR and plotted as percentage of mRNA detected in the controls treated with the corresponding transfection reagent alone. (B) Time course of P2X₃ protein inhibition by 100 nM siRNA-7556/7557 versus its mismatch analogue siRNA-MM-7558/7559. Western blotting followed by P2X₃-specific immunodetection revealed the expression levels shown below (an average value from two experiments). Time points as indicated at the top. Molecular weights of two glycosylated forms of P2X₃ are shown on the left.

RNAi-mediated silencing of P2X₃ expression by siRNA-7556/7557 was effective but not complete. Analysis of potential contributing factors included the targeted sequence considered in siRNA design, the type of 2 nt 3′-overhang modification and the age of the pretreated cell culture. Four siRNA duplexes were tested: (i) siRNA-7556/7557 to target an optimal sequence, i.e. a sequence available for antisense interaction and equipped with noncomplementary, MOE-modified 3′-overhangs; (ii) siRNA-4882/4883, similar to the previous siRNA but equipped with dT-modified 3′-overhangs; (iii) siRNA-8646/8647, again a similar reagent but equipped with fully complementary MOE-modified 3′-overhangs; and (iv) siRNA-4884/4885 to target a different sequence selected according to T. Tuschl’s rules (49) with complementary dT-modified 3′-overhangs (Table 1). All four siRNAs showed high silencing potential in CHO-rP2X₃ cells grown up to 70% confluency before seeding for transfection (Fig. 3A). Cultures in the exponential growth phase seem to respond more effectively to RNAi; up to 95% P2X₃ mRNA inhibition was observed with the application of 200 nM siRNA reagents 24 h after transfection. The siRNA-8646/8647 duplex was selected for further experiments. Its silencing potential was confirmed on the protein level by western blotting and immunodetection (Fig. 3B).

Figure 3.

Sequence-specific siRNA-mediated repression of P2X₃. (A) P2X₃ mRNA inhibition by 200 nM siRNA duplexes. Twenty-four hours after transfection of CHO-rP2X₃ cells, P2X₃-specific mRNA was measured with Q-PCR and plotted as percentage of mRNA detected in the control treated with Oligofectamine alone. Sequences and modifications are shown in Figure 1B and Table 1. (B) P2X₃ protein reduction by 200 nM siRNA-8646/8647, but not by its mismatch analogue siRNA-MM-7558/7559 or the unrelated siRNA-7126/7127. Twenty-four hours after transfection, protein was extracted and analysed by western blotting. P2X₃-specific immunodetection reveals expression levels as shown below (an average value from two experiments). Time points as indicated at the top. Molecular weights of two glycosylated forms of P2X₃ are shown on the left.

Using an optimal ASO (19) and siRNA, we screened for a convenient transfection method allowing efficient delivery of both reagents. A standard electroporation procedure was selected for the transfection of the CHO-rP2X₂/P2X₃ cell line, expressing both P2X₂ and P2X₃ proteins. Use of this particular cell line allows monitoring of P2X₃ protein expression via a functional assay (FLIPR), which requires both proteins to achieve intense signalling (G. Dorn, unpublished results). Similar to other tested cell lines, electroporation of siRNA-8646/8647 in CHO-rP2X₂/P2X₃ resulted in dose-dependent silencing of P2X₃ expression at the mRNA and protein levels, as demonstrated by Q-PCR and FLIPR assay (Fig. 4A). P2X₃ mRNA inhibition reached 60–70% with 200 nM Oligofectamine-transfected (Fig. 2A) and 50 nM electroporated siRNA (Fig. 4A), indicating much more efficient delivery with the latter method. Interestingly, siRNA-mediated silencing reached a plateau above 50 nM of electroporated siRNA. This tendency persisted during the time-course experiment (Fig. 4B).

Figure 4.

P2X₃ expression in electroporated CHO-rP2X₂/P2X₃ cells. (A) Dose-dependent inhibition by siRNA-8646/864 versus its mismatch analogue siRNA-MM-7558/7559. Concentrations of siRNAs as indicated. Twenty-four hours after electroporation, P2X₃-specific mRNA was measured with Q-PCR and plotted as percentage of mock electroporated sample (white bars, untreated). Forty-eight hours after electroporation, cells were subjected to the functional assay (Ca-influx measured by FLIPR; black bars). (B) Time-course of P2X₃ mRNA inhibition by four doses of siRNA-8646/8647 versus its mismatch analogue siRNA-MM-7558/7559 and its antisense analogue ASO-5037. mRNA was measured with Q-PCR and plotted as percentage of untreated samples collected at the relevant time points as indicated.

In order to analyse potential interaction between the RNAi and the antisense approach, a series of titration experi ments was performed. A suboptimal concentration of 50 nM siRNA-8646/8647 was chosen (Fig. 4B) as a background for titration with increasing concentrations (10, 50, 100 or 200 nM) of three ASO reagents: (i) a homologous ASO-5037; (ii) a nonhomologous ASO-5038; and (iii) ASO-MM-5655, a mismatch control. When electroporated separately, both siRNA-8646/8647 and its homologue ASO-5037 downregulated P2X₃ mRNA in a dose-dependent manner (Fig. 5A). In the titration experiment, 200 nM ASO-5037 had no additional effect on the RNAi-mediated background inhibition of P2X₃ mRNA, illustrating the greater potency of siRNA than homologous ASO. A similar effect of the homologous ASO on RNAi was detected at the mRNA level by Q-PCR (Fig. 5A) and at the functional level, as demonstrated by FLIPR (Fig. 6A). At low concentrations of 10–50 nM, ASO-5037 slightly enhanced P2X₃ mRNA inhibition in cooperation with siRNA-8646/8647. This effect, however, does not seem to be sequence-specific as it was observed only occasionally with different ASO reagents, including the mismatch ASO- MM-5655 (see below). In the next step, combinations of siRNA-8646/8647 and nonhomologous ASO-5038 were analysed. Electroporated separately, both reagents successfully downregulate P2X₃ mRNA in a dose-dependent manner (Fig. 5A and B, respectively). Combination of the two reagents, using 50 nM siRNA and increasing concentrations of ASO-5038 (Fig. 5B) enhanced P2X₃ mRNA inhibition. For example, 50 nM siRNA and 200 nM ASO-5038 provoked 90% inhibition, while 50 nM siRNA-8646/8647 or 200 nM ASO-5038 alone affected 60 and 70% inhibition of P2X₃ mRNA, respectively. As expected, mismatch ASO-MM-5655 did not affect P2X₃ mRNA inhibition when used alone, nor did it affect RNAi when combined with siRNA-8646/8647 (Fig. 5B). In summary, a potent ASO can, together with siRNA-8646/8647, cooperatively inhibit P2X₃ expression if it targets a distant sequence (e.g. ASO-5038).

Figure 5.

Direct comparison of P2X₃ targeting by ASO and siRNA variants in electroporated CHO-rP2X₂/P2X₃ cells analysed by Q-PCR. ASO or siRNA (10, 50, 100 and 200 nM) are compared to 200 nM mismatch controls (siRNA-MM-7558/7559, ASO-MM-5655 and ASO-5596). Alternatively, samples were treated with a mixture of 50 nM siRNA and increasing concentrations of ASO (10, 50, 100 or 200 nM), as indicated. At 24 h after electroporation, P2X₃-specific mRNA was measured with Q-PCR and plotted as percentage of mock electroporation (untreated). For clarity of presentation, samples are grouped in logical series and shown with a consistent colour (white or black). (A) Dose-dependent P2X₃ mRNA inhibition by siRNA-8646/8647, an homologous ASO-5037, and by a mixture of the two (H-mix). (B) Dose-dependent P2X₃ mRNA inhibition by siRNA-8646/8647 is unaffected by titration with ASO-MM-5655 (MM-mix), but enhanced by nonhomologous ASO-5038 (NH-mix). (C) Dose-dependent P2X₃ mRNA inhibition by siRNA-4884/4885 is not sensitive to the homologous ASO-10326 (H-mix), but enhanced by nonhomologous ASO-5038 (NH-mix).

Figure 6.

Direct comparison of P2X₃ targeting by ASO and siRNA variants in electroporated CHO-rP2X₂/P2X₃ cells measured by the FLIPR functional assay. ASO or siRNA (10, 50, 100 and 200 nM) are compared to 200 nM mismatch controls (siRNA-MM-7558/7559, ASO-MM-5655 and ASO-5596). Alternatively, samples were treated with a mixture of 50 nM siRNA and increasing concentrations of ASO (10, 50, 100 or 200 nM), as indicated. At 48 h after electroporation, cells were subjected to a functional assay (Ca-influx measured by FLIPR) and plotted as percentage of mock electroporation (untreated). For clarity of presentation, samples are grouped in logical series and shown with a consistent colour (white or black). (A) Dose-dependent inhibition of the functional response (Ca-influx) to 10 µM of the agonist α,β-methylene-ATP by transfection with siRNA-8646/8647, an homologous ASO-5037 and by a mixture of the two (H-mix). (B) Dose-dependent inhibition of the functional response (Ca-influx) to 10 µM α,β-methylene-ATP by transfection with siRNA-4884/4885, an homologous ASO-10326 and a mixture of the two (H-mix). Note that dose-dependent P2X₃ receptor inhibition by siRNA-4884/4885 is not sensitive to the homologous ASO-10326 (H-mix), but enhanced by nonhomologous ASO-5037 (NH-mix).

The observed combinatorial effect between siRNA-8646/8647 and nonhomologous ASO-5038 targeting P2X₃ mRNA may have a more general character. To analyse this possibility, we performed a series of titration assays using another siRNA-4884/4885, which targets the AUG-proximal region in the coding sequence of the P2X₃ mRNA (Fig. 1A). When electroporated separately, siRNA-4884/4885 and its homologue ASO-10326 downregulated P2X₃ mRNA in a dose-dependent manner (Fig. 5C). In the titration experiment, 200 nM ASO-10326 had no obvious effect on the RNAi-mediated background inhibition of P2X₃ expression (Figs 5C and 6B). However, RNAi-mediated silencing of the P2X₃ receptor was again enhanced by increasing concentrations of nonhomologous ASO-5038 (Fig. 5C). Furthermore, another combination of siRNA-4884/4885 and ASO-5037, which in this experiment played the role of a nonhomologous ASO, resulted in enhanced P2X₃ inhibition (Fig. 6B). These results confirm the independent coordinated effect of ASO and siRNA when the two reagents target nonoverlapping sequences on a common mRNA.

DISCUSSION

The rat ligand-gated ion channel P2X₃ is attracting increasing attention as a model receptor proposed and now proven to be involved in pain signalling (6). In the absence of selective pharmacological ligands for the P2X₃ receptor, an antisense approach was recently employed in vitro (19) and in vivo (20) to specifically inhibit P2X₃ gene expression and to study its function. ASOs have been gaining increasing acceptance for investigating fundamental problems in neurobiology and for studying and possibly ameliorating or treating neurological disorders (18). However, the well documented non-specific effects of ASO treatment complicate interpretation of studies employing antisense technology (21). As an alternative to ASOs, sequence-specific degradation of mRNAs can also be triggered by the RNAi process. In the present work, we designed and analysed in vitro a series of siRNA duplexes that specifically inhibit expression of the rat recombinant P2X₃ receptor. Synthetic siRNAs against the P2X₃ receptor were delivered transiently by a single cation lipid-mediated transfection or electroporation. In our system, dramatic differences in lipid-mediated transfections were observed, depending on the chemistry of the transfection reagent. Oligofectamine was optimal with siRNAs but not for ASO transfections; conversely, Lipofectin worked well with ASO (19) but not with siRNA duplexes. Eventually, standard electroporation was chosen for the effective delivery of both ASO and siRNA, enabling comparison of the antisense and RNAi approaches. Additionally, the similarity of direct administration of nucleic acids during electroporation in vitro and intrathecal injection in vivo (20) implies that transfection by electroporation is especially valid for the initial screening of ASO and siRNA reagents, prior to their administration in animal model systems.

Oligofectamine-mediated transfection of 200 nM siRNA duplex resulted in 60–90% downregulation of P2X₃ mRNA, depending on the cell line (Fig. 2A: cell line 33B and CHO-K1, respectively) and the confluency of the culture prior to transfection. Similar to Harborth et al. (41), we recommend the use of subconfluent cultures for seeding cells, 24 h before transfection. In any case, a similar range of inhibition was usually observed with 50–100 nM siRNAs transfected by electroporation, implying a greater efficiency of delivery with the latter method. During our study, we compared the potential of three ASO and four siRNA reagents. Both ASO-5037 and -5038 were selected earlier as very potent P2X₃ inhibitors, thought to target accessible regions in the P2X₃ mRNA structure (19). In the current work, ASO-5037 performed less efficiently than the siRNA-8646/8647 reagent with the overlapping sequence. All ASO used in this work are heavily modified with MOE ribonucleosides (Table 1), which might affect the efficiency of the electroporation process per se. In fact, siRNA duplexes are modified only at the 2 nt 3′-overhangs, which may result in more efficient electroporation and in general higher performance (Fig. 5). Three of the siRNAs tested take advantage of the locally available mRNA structure and, thus, were designed to target an accessible sequence covered by ASO-5037 (19). siRNA-4884/4885 targets an AUG-proximal region selected ad hoc covered by suboptimal ASO-10326. Despite being targeted to regions far apart in the P2X₃ coding sequence (Fig. 1), all four analysed siRNAs performed efficiently, resulting in 90–95% inhibition of P2X₃ mRNA (Fig. 3A). Our results imply that an RNAi-based method for targeting gene expression is less sensitive to the local restraints of the mRNA structure than is antisense. This subject has been already addressed by other researchers (32,40,41,50,51).

Similar to others (50,52), we found that siRNA-mediated P2X₃ mRNA silencing was transient and, in our experimental conditions, did not last much longer than 72 h, as monitored by Q-PCR at the mRNA level (Fig. 4B). These results thus suggest the absence of a propagative, RdRP-dependent system (reviewed in 53) for RNAi in rodents. Although efficient, the inhibition of P2X₃ gene expression by siRNAs was not complete and might result from the relatively high expression of the recombinant P2X₃ gene in all three tested cell lines. In natural circumstances, the P2X₃ receptor expression is restricted to particular sensory neurons and is upregulated following neuropathic injury. To study the downregulation of P2X₃ expression, we made use of stably transfected cell lines, specifically selected and expressing a high level of recombinant rat P2X₃ receptor driven by the β-actin promoter. Due to the nature of the P2X₃ target, general strong immunostaining of the P2X₃ protein in the recombinant cell lines (G. Dorn, unpublished results) or strong inhibition of its expression have no apparent side-effects or further consequences for cell biology and survival. Thus, in our system, the results of ASO- or siRNA-mediated inhibition of the P2X₃ expression can be interpreted simply in terms of the performance of the employed reagents.

The question of how to ameliorate performance of the RNAi method is being addressed in many laboratories using RNAi as a simple method of reverse genetics to study consequences of ‘knock-out’ and ‘knock-down’ of gene expression. To be comparable with classical genetic deletions, RNAi should lead to the complete and stable degradation of the targeted mRNA. From practical laboratory experience, it is known that longer lasting effects of RNAi can be achieved in vitro by a two-step transfection with a single siRNA reagent or by intracellular expression of siRNA from the plasmid DNA (reviewed in 47). However, results from cell-transfection methods do not easily apply to the in vivo chronic pain models, where high doses of ASO reagent have been directly and continuously (for more than 7 days) administered by intrathecal injection (20). Results of in vivo experiments involving siRNA to target P2X₃ receptor have not been published yet but, if the conditions of siRNA administration are similar to those of ASO, then the estimated cost of the siRNA reagent will be substantially higher. One way to lower the cost of in vivo siRNA therapies might be to use mixed siRNA reagents, if combinations of siRNAs result in increased activity through synergism or addition. Tests of this concept by co-transfecting various combinations of multiple siRNAs revealed no addition nor synergism (50,51). On the contrary, co-transfection with multiple siRNAs might reduce the activity of the most potent form, probably due to sequence-independent competition for enzymatic components of the RNAi process (‘diluting effects’) (41).

In the present study, an alternative approach was tested using combinations of ASO and siRNA reagents which are known to stimulate mRNA degradation via activation of different biological processes in the cell. There is evidence that siRNA degrades only mature mRNAs by activating RNAi process in the cytoplasm, but antisense agent can affect both pre-mRNAs and mature transcripts by activating an ubiquitous RNaseH. Consequently, simultaneous application of both reagents might enhance mRNA degradation. To test this hypothesis, we took advantage of the P2X₃ receptor, an important target in chronic pain, which has been extensively studied in vitro (19) and in vivo (20) using ASOs. During a series of titration experiments with various combinations of siRNA and ASO reagents, we indeed observed enhanced inhibition of P2X₃ when siRNA was combined with the nonhomologous ASO reagent. This combinatorial action might be a general phenomena, as the interaction can occur between different pairs of siRNA and potent nonhomologous ASO targeting common mRNA. The observed effect is definitely sequence-dependent and does not occur if siRNA is combined with a mismatched control or homologous ASO. The latter probably competes with the cytoplasmic RISC complex for exactly the same location on the mRNA. Although this is justified by the overlapping occurrence of RNAi (54) and RnaseH (55,56) in the cytoplasm, the molecular mechanism of the competition between both processes remains elusive.

In summary, we have detected and characterised an independent combinatorial action of ASOs and RNAi- mediated specific inhibition of P2X₃ receptor. Enhanced downregulation was observed when siRNA was combined with nonhomologous ASO, targeting distant regions on the common P2X₃ mRNA. When tested in vitro, the cooperation between nonhomologous ASO and siRNA was persistent but not dramatic. Combination of the two reagents allows more efficient downregulation of P2X₃ mRNA than separate application, as was observed at both the mRNA and protein levels. Based on the rich literature describing in vivo antisense studies, we suppose that even modest effects on expression of a specific protein can have dramatic behavioural effects in vivo. Further developments with animal models will show whether humans could profit one day from mixed siRNA and ASO therapy, in which combinations of reagents at low dose may result in high and specific efficacy with low side-effects and offer an improved therapeutic window.