Specific inhibition of pathological prion protein accumulation by small interfering RNAs

Transmissible spongiform encephalopathies (TSEs), also named `prion diseases', are fatal neurodegenerative diseases including Creutzfeldt-Jakob disease in humans, scrapie in sheep and goats, and bovine spongiform encephalopathy in cattle (Prusiner et al.,1996; Caughey and Chesebro,1997). In affected humans and animals, TSEs are characterized by the accumulation in the central nervous system of a protease-resistant isoform(PrP-res or PrPsc) of the host-encoded cellular prion protein (PrP-sen or PrPc). As no differences in their primary sequences could be demonstrated,PrP-sen and PrP-res isoforms are believed to differ only in their tertiary structure (Stahl et al.,1993). Macromolecular aggregates of PrP-res are thought to continuously grow by binding PrP-sen molecules and converting them into their own conformation. Conversion of PrP-sen to PrP-res is the pivotal event in the etiology of TSEs, even though the mechanisms initiating the conversion reaction remain unknown. Both PrP isoforms, PrP-sen and PrP-res, are believed to be involved in the neuropathogenesis of TSEs(Büeler et al., 1993; Brandner et al., 1996). Homozygous disruption of the Prnp gene encoding PrP results in a complete scrapie resistance in these mice(Büeler et al., 1993). Moreover, the expression level of PrP-sen is a major factor influencing the rate of replication of the infectious agent as well as the incubation time of the TSEs (Sakaguchi et al.,1995; Carlson et al.,1994). The importance of PrP-sen in TSEs is also exemplified by genetic linkages between familial TSEs and mutations in the human Prnp gene (Parchi and Gambetti,1995).

Chronically scrapie-infected cell lines have been extensively used as models for selecting `anti-prion' drugs such as porphyrines(Caughey et al., 1998),amphotericin B (Mange et al.,2000) and quinacrine (Doh-Ura et al., 2000). Since PrP-sen expression is essential for both TSE pathogenesis and conversion leading to PrP-res accumulation, we studied the effect of specific Prnp gene silencing molecules in scrapie-infected neuroblastoma cells (N2aS12sc⁺). Small interfering RNAs (siRNAs)provide new powerful tools to potentially silence any targeted gene(Elbashir et al., 2001). Because siRNA duplexes trigger specific gene silencing in mammalian somatic cells without the activation of any unspecific response, the analysis of gene function in cultured cells has now become possible. Here, we show that transient transfection of siRNA duplexes designed from the mouse Prnpgene (moPrnp) causes a rapid loss of PrP-sen expression and abrogates PrP-res formation in N2aS12sc⁺ cells. By contrast, the scrambled siRNA failed to inhibit both the PrP-sen expression and PrP-res accumulation in these cells suggesting a highly specific siRNA effect. Finally, we demonstrate that the siRNA gene silencing is independent of both the host cell types and mouse-adapted scrapie strain.

Proteinase K (PK) and Pefabloc were purchased from Boehringer Mannheim. Dulbecco's modified Eagle's medium (DMEM), Opti-MEM, trypsin, geneticin G418,and Oligofectamine™ were from Invitrogen, and fœtal bovine serum(FBS) was from BioWest (Nuaille, France). Mouse monoclonal antibodies SAF83 and SAF84 were raised in knock-out mice(Manson et al., 1994) by immunization with a scrapie-associated fibrils (SAF) preparation obtained from hamster-infected brain (263K strain). SAFs were denatured before immunization by treatment with formic acid. Both antibodies crossreact with PrP from most mammalian species (bovine, ovine, mouse, hamster and human). Secondary antibodies conjugated to peroxidase were from Jackson ImmunoResearch (West Grove, PA).

Neuroblastoma N2a cells over-expressing mouse PrP (subclone #58) were cultured in Opti-MEM containing 10% inactivated FBS (FBSi),penicillin-streptomycin and 1 mg/l G418. This cell line named N2aS12sc⁺ has been chronically infected with brain homogenates of Chandler strain-infected mice (Nishida et al., 2000). Since the treatment of N2aS12sc⁺ cells with Congo Red (1 μg/ml) totally cured the cells of PrP-res, Congo Red-treated N2aS12 cells were used as uninfected cells (N2aS12). The GT1 cells, infected with brain homogenates of 22L murine scrapie strain, were grown in Dulbecco modified Eagle's medium containing 5% fetal calf serum, 5% horse serum, and penicillin-streptomycin. All cultured cells were maintained at 37°C in 5%CO₂ and split 1:4 every 4 days.

siRNAs corresponding to the moPrnp gene from codon 392 to 410 were synthesized by Eurogentec (Seraing, Belgium). Typically, siRNAs were made of 19 ribonucleotides followed by two extra thymidine bases at the 3′ end overhang on both strands. The specific siRNAs sequences used were:5′-GCC-CAG-CAA-ACC-AAA-AAC-CTT-3′ (sense) and scramble 5′-CGC-ACC-AGA-ACA-AAC-ACA-CTT-3′ (sense). The specific siRNA sense 5′-GCC-CAG-CAA-ACC-AAA-AAC-CTT-3′ was labeled with 6-carboxyfluorescein (6-FAM) on the 3′ end. Annealing for duplexes siRNA formation was performed by incubation at 92°C for 1 minute in 50 mM Tris-HCl pH 7.5 containing 100 mM NaCl, followed by a 60 minute incubation at room temperature. The final stock concentration of siRNA duplexes (20 μM)was stored at 4°C until use.

Twenty-four hours before the transfection, cells were seeded at 10-15%confluence in a 12-well culture plate with appropriate culture medium. Varying amounts of double strand siRNA (1-10 μl) were mixed with the corresponding half-volume (0.5-5 μl) of Oligofectamine™ reagent for 20 minutes according to the manufacturer's instructions. The mixture was then applied to the cells in a final volume made up to 500 μl with Opti-MEM without FBSi and antibiotic. After incubation for 4 hours at 37°C under 5%CO₂, 250 μl of Opti-MEM supplemented with 30% FBSi and a penicillin/streptomycin mixture were added. Cells were then cultured for three days at 37°C until confluent.

Confluent cultures were lysed for 10 minutes at 4°C in lysis buffer (50 mM Tris-HCl pH 7.4 containing 150 mM NaCl, 0.5% Triton X-100, 0.5% sodium deoxycholate, 5 mM EDTA), then centrifuged for 5 minutes at 3000 g. For detection of PrP-sen, one-tenth of post-nuclear supernatants were directly mixed with the same volume of 2× denaturing loading buffer. For detection of PrP-res, samples were digested with 20 μg of PK per mg of total proteins for 30 minutes at 37°C. The digestion was stopped by adding Pefabloc (1 mM) for 5 minutes before centrifugation at 20,000 g for 90 minutes at room temperature. Pellets were resuspended in 30 μl of denaturing loading buffer, sonicated, boiled for 5 minutes and loaded onto a 12% polyacrylamide gel. Proteins were separated by SDS-PAGE, then electroblotted onto a nitrocellulose membrane (Protran BA83,Schleicher & Schuell). Membranes were treated with a solution containing 5% nonfat dried milk in 20 mM Tris-HCl pH 8, 100 mM NaCl, 0.1% Tween 20, and incubated overnight at 4°C with the appropriate primary antibody. Blots were developed by using an enhanced chemiluminescence system (Pierce,Rockford, IL) and exposed on X-ray films (X-OMAT AR, Kodak). Densitometry was performed with `National Institutes of Health' Image software, on at least four independent experiments and results expressed as a percentage of control levels.

N2aS12 cells were grown on glass coverslips (Lab-Tek) in Opti-MEM supplemented with 10% FBSi. Cells were transfected as described above with the 6-FAM-conjugated siRNAs duplexes (400 nM) or the single-stranded 6-FAM-conjugated siRNA as a control experiment. At the indicated post-transfection times, coverslips were washed twice with phosphate-buffered saline (PBS) and fixed in paraformaldehyde 3% in PBS for 10 minutes at room temperature. Cells were washed twice with PBS and then incubated in PBS containing 50 mM NH₄Cl for 10 minutes in order to quench excess free aldehyde groups. Cells were permeabilized in PBS containing 0.1% Triton X-100 and 1 μg/ml propidium iodide for 10 minutes at room temperature. Coverslips were washed twice with PBS, and then mounted on glass slides with mowiol. Cells were analysed using a laser scanning confocal inverted microscope (Leica, TCS-SP) equipped with an argon-krypton laser. Samples were scanned under both 488 nm and 568 nm excitation wavelength for 6-FAM-conjugated siRNA and for propidium iodide, respectively. Images were acquired as single transcellular optical sections and averaged over at least 8 scans per frame.

It has been previously shown that the relative level of PrP-res production correlates with the level of PrP-sen expression. These observations led us to study the effect of silencing the Prnp gene using siRNAs. We speculated that abrogation of PrP-sen expression may subsequently downregulate, or abolish PrP-res accumulation in scrapie-infected N2aS12sc⁺ cells. These cells were transfected with various concentrations of moPrnp gene-specific siRNA and then assayed for PrP immunoreactivity in the absence (Fig. 1A) or in the presence (Fig. 1B) of PK-digestion. The expression of PrP-sen was specifically reduced by the cognate siRNA duplex in a dose-dependent manner, whereas expression of the unrelated protein ERK was unaffected(Fig. 1A). Importantly, this treatment also reduced PrP-res levels (Fig. 1B). The concentrations of Prnp-specific siRNA able to inhibit 50% of the PrP-sen expression and of the PrP-res accumulation were IC₅₀ ∼72 and 100 nM, respectively(Fig. 1C). Scrambled siRNA unrelated to any known genes, was unable to perturb the amount of PrP-sen and PrP-res in N2aS12sc⁺ cells even at high concentrations (400 nM). Single-stranded siRNAs transfected independently in N2aS12sc⁺ cells were also unable to diminish the expression of PrP-sen (data not shown).

To assess whether the nature of the host cell or the scrapie strain could influence the siRNA inhibition effect, siRNAs were transfected into the mouse hypothalamic cell line (GT-1) chronically infected with the mouse scrapie strain 22L (Nishida et al.,2000). The detectable amounts of PrP-res were also drastically reduced in this cell line after transfection with Prnpgene-specific siRNA duplexes (Fig. 2) and IC₅₀ values were comparable with those obtained using N2aS12sc⁺ cells (IC₅₀ ∼106 nM). This suggests that neither host cell types nor mouse scrapie strains influence the gene-silencing effect of the siRNA. During the time course of these experiments (4 days), no cytotoxic effect of siRNAs was observed.

To estimate the efficiency of siRNA transfections into the neuroblastoma cells, confocal microscopy analysis was performed using fluorescein-labeled siRNA duplexes (Fig. 3). The percentage of fluorescein-labeled neuroblastoma cells (green) reached∼82%±6% as early as 48 hours post-transfection and was stable over 72 hours (Fig. 3A). No green fluorescence was detected in cells transfected with the fluorescein-labeled single-stranded RNA as shown in the control experiment(Fig. 3A). We also used confocal imaging to investigate the sub-cellular localization of the fluorescein-labeled siRNA duplexes. Transfected cells exhibited a punctate fluorescence pattern largely localised to the perinuclear cytoplasmic region(Fig. 3B). No labeling was observed at the cell surface or inside the nucleus. Based on both the transfection efficiency and the assessment of PrP expression, we found that Oligofectamine™ was the most efficient transfection reagent for siRNA duplexes.

To determine whether the PrP-res depletion induced by siRNA was transient or permanent, chronically infected N2aS12sc⁺ cells were transfected once with 400 nM siRNA and then cultured for either one and two weeks through two and four passages, respectively. Samples were assayed by Western blot analysis for the presence of PrP isoforms in the presence or in the absence of PK-digestion (Fig. 4). PrP-sen was barely detectable in transfected cells before passaging but levels increased with the number of passages, reaching the level observed with untreated-cells after four passages. A single siRNA transfection led to a complete disappearance of PrP-res immunoreactivity after two passages of N2aS12sc⁺ cells. However, after four passages, low levels of the PK-resistant PrP isoform were observed in N2aS12sc⁺ cells. Our data suggest that a single exposure of N2aS12sc⁺ cells to siRNA was not sufficient to completely remove the pathogenic isoform and permanently suppress scrapie infection.

In the present report, we demonstrate that abrogation of PrP-sen synthesis by siRNA-mediated suppression of the Prnp gene inhibits the formation of PrP-res in chronically scrapie-infected cells. In neuroblastoma cells,PrP-mRNAs have a half-life of approximately 7 hours(Pfeifer et al., 1993) and PrP-sen is synthesized and degraded relatively rapidly (t_g ∼5 hours). Whereas, PrP-res aggregates which derived from PrP-sen, is produced slowly (t_g ∼15 hours)(Borchelt et al., 1990). Several mechanisms may lead to PrP-res depletion including interference with the PrP-sen/PrP-res interaction (Caspi et al., 1998; Chabry et al.,1998; Horiuchi et al.,2001) and reversion of PrP-res to a PK-sensitive state(Soto et al., 2000). In addition, prevention of PrP-sen endocytosis from the cell surface to endosomes and/or release of PrP-sen from the plasma membrane could prevent PrP-res accumulation (Enari et al.,2001; Marella et al.,2002). Expression of PrP-sen is essential but not sufficient for prion propagation and pathogenesis (Raeber et al., 1999). It has been suggested earlier that suppression of PrP expression might be a potential therapeutic approach against prion diseases (Büeler et al.,1993).

The factors responsible for the rate and the strength of siRNA-induced gene suppression are currently unknown. Here, we demonstrate that siRNA designed from the moPrnp gene was highly specific to cells expressing the PrP mouse sequence. The siRNA effect was independent of both the scrapie strains and the host cell type. Furthermore, no difference in siRNA efficiency was observed between moPrP overexpressing cells (N2aS12 subclone) and N2a cells expressing only endogenous PrP (data not shown). The finding that exposure of chronically infected neuroblastoma cells to siRNA results in undetectable amounts of PrP-res after 3-4 days, suggests that these cells express proteases that degrade PrP-res aggregates. Recently, Enari et al. have proposed that the cellular level of PrP-res is determined by the rate of its formation from the substrate, PrP-sen, and its catabolism(Enari et al., 2001). The removal of PrP-sen could displace the equilibrium to the degradation processes. An obvious limitation is that although PrP-res levels were dramatically decreased by acute siRNA exposure, the cells were not definitively `cured' of scrapie infection; thus, chronic treatments or longer exposures of siRNA may be considered in the future. We are currently performing experiments to improve the long-term effects of siRNA duplexes.

Interestingly, a recent report has shown the selective suppression of gene expression induced by siRNA in primary mammalian neurons(Krichevsky and Kosik, 2002). The role of PrP and its interaction with putative partners in neurons have not yet been clarified. Thus, siRNA technology should provide insights into the understanding of the physiological functions of PrP in somatic cultured cells.

Although there is likely to be technical difficulties with siRNA delivery in vivo, its use to inhibit gene expression in mice has recently been reported(Lewis et al., 2002). Although efficient and powerful brain delivery systems remain to be found, it appears likely that siRNA technology could constitute a new promising strategy for therapy against TSEs. Current drug treatments delay, but do not prevent, the appearance of clinical symptoms and death in experimental scrapie-infected animals (Demaimay et al., 1997; Priola et al., 2000). The in vivo use of siRNA alone, or in combination with these drugs, should be considered to improve the life span and cure sick animals. The recent development of siRNA technology offers a powerful tool to ablate specific gene expression in a wide range of mammalian cells including neurons.

Our results show that the siRNA technique should be useful for dissecting basic mechanisms of prion pathogenesis, in addition to its obvious potential as a therapeutic agent. Ultimately it may give effective results in therapy of numerous neurodegenerative diseases linked to expression of misfolded endogenous proteins such as Alzheimer's, Huntington's and Parkinson's diseases.

We are grateful to J. Grassi and S. Lehmann for providing us with the monoclonal antibodies SAFs and the N2aS12sc⁺and GT-1sc⁺cells, respectively. We thank N. Zsürger for help with the preparation of the figures and R. Pichot for technical assistance. Very special thanks are due to M. Borsotto, D. Vilette, G. Walker and J.-P. Vincent for critical reading of the manuscript. This work was supported by the Centre National de la Recherche Scientifique, the Action Concertée Incitative Jeunes Chercheurs (2000) and the Groupement d'Intérêt Scientifique:Infections à Prions (2001).