Small interfering RNA (siRNA) is an extremely effective tool for reducing gene expression in a variety of systems, including mammalian cells. Gene specific siRNAs have typically been made by chemical synthesis. Producing them by chemical synthesis, however, is costly, and while most siRNAs exhibit some effect, only about 25% of those designed reduce target levels by 80% or more (1). A new in vitro procedure that cleaves long dsRNA into a mixture of siRNAs significantly reduces the cost of siRNA preparation. The resulting siRNA population is effective at mediating gene specific silencing and bypasses the need to screen for functional siRNAs.
In the RNA interference (RNAi) pathway in C. elegans, Drosophila, and other diverse organisms, long double-stranded RNA (dsRNA) introduced into a cell is cleaved into a mixture of siRNAs by an RNase III-like endoribonuclease (Dicer) producing siRNAs that are 21-23 bp in length and contain 3' dinucleotide overhangs with 5'-phosphate and 3'-hydroxyl termini. These siRNAs mediate degradation of homologous transcripts, leading to gene silencing (2-5; see also "Gene Specifi c Silencing by RNAi").
Escherichia coli RNase III, which is involved in the maturation and degradation of diverse cellular, phage, and plasmid RNAs (6-9), digests long dsRNA to short duplex products ranging from ~12 to 15 bp in length with termini identical to those produced by Dicer (9). The average product length generated by RNase III digestion can be increased by altering digestion conditions, and it has recently been shown that these longer products (21 bp range) can effectively mediate RNAi in both mammalian cells (10) and mouse embryos (11). We demonstrate here that the siRNA cocktails produced by complete digestion by RNase III (~12 to 15 bp) are capable of silencing specific genes at levels comparable to chemically or enzymatically synthesized siRNAs.
Efficient Digestion of Distinct dsRNA Sequences. Using optimized digestion conditions we analyzed the ability of RNase III to digest a number of long dsRNA substrates. Human GAPDH, La, and c-FOS dsRNA (200 bp) was prepared by in vitro transcription (Silencer siRNA Cocktail Kit (RNase III); See Materials and Methods). The dsRNA was digested using 1 U RNase III per microgram of RNA for 1 hour at 37C, to generate siRNA cocktails for each target gene. After a 1 hour digestion with RNase III, the long dsRNAs were reduced to fragments <30 bp, with the majority between 1215 bp (Figure 1A). In addition, dsRNAs to Cyclophillin, c-myc, Map Kinase 9, PKC-alpha, Raf-1, Nautilus, and h-ras made as described above, were also digested with similar results (data not shown). This demonstrates the ability of the bacterial RNase III enzyme to efficiently digest a variety of dsRNA sequences.
Silencing by RNase III Digested dsRNA. We next analyzed the silencing ability of the RNase III generated siRNA cocktails. GAPDH and La proteins in HeLa cells are abundant and endogenous levels are easily detected. However the endogenous level of c-FOS in 293 cells is relatively low, and reduction in protein levels makes the protein undetectable. In order to overcome this limitation, 293 cells were stimulated to increase c-FOS protein levels by the addition of 50 nM phorbol ester (PMA) for 24 hours prior to protein analysis. RNase III-generated siRNA cocktails to GAPDH and La were transfected into HeLa cells, and the c-fos siRNA population was transfected into 293 cells. Immunofluorescence was used to examine the gene silencing effect. The fluorescent signal from this experiment was then quantitated and normalized for cell number. Figure 1B shows that protein levels were reduced by 78% for GAPDH, 86% for La, and 75% for c-FOS by introduction of the respective siRNA cocktails. These data demonstrate that RNase III generated siRNAs are very efficient at reducing target gene expression.
Figure 1. Silencing Gene Targets by RNase III Derived siRNA Cocktails. A 200 bp dsRNA (15 g) for each gene of interest was digested with 2.5 U RNase III for 1 hour at 37C. 1A. RNase III efficiently digests dsRNA. One microgram of the dsRNA before and after RNase III dig estion was run on a 15% non-denaturing acrylamide gel along with a 21 bp chemically synthesized siRNA to GAPDH, which served as a size marker. The gel was stained with ethidium bromide and photographed under UV light. 1B. RNase III derived siRNA cocktails silence GAPDH, La and c-FOS. GAPDH and La siRNA cocktails were transfected into HeLa cells. The c-fos siRNA mixture was transfected into 293 cells followed by 24 hours of stimulation with 50 nM PMA. All samples were harvested at 48 hours post transfection and immunofluorescence was performed with the appropriate antibodies. Fluorescence signal was quantitated, normalized for cell number and graphed.
Silencing by 12-15 bp RNase III Digestion Products. The size of chemically synthesized siRNA most often used for mediating RNAi is 21 bp (12). It has been shown that the 21 bp products generated by RNase III digestion are potent inhibitors of gene expression (10). However the products of a complete RNase III digestion are 12-15 bp. To compare the ability of these smaller products to reduce gene expression with 21 bp siRNA, a 200 bp GAPDH dsRNA was digested with RNase III under standard conditions and the resulting 12-15 bp fragments were acrylamide gel purified from the incomplete digestion products (Figure 2A). HeLa cells were transfected with 100 nM final concentration of the 12-15 bp purified products, as well as with the same concentration of a 21 bp chemically synthesized siRNA known to effectively reduce GAPDH levels. Figure 2B shows representative immunofluorescence images demonstrating reduction in GAPDH levels after transfection with the RNase III generated siRNAs. The 12-15 bp products are capable of reducing targ et gene expression at comparable levels to a chemically synthesized siRNA targeting GAPDH (Figure 2C). This experiment demonstrates that the smaller sized siRNA cocktails produced by RNase III reduce target gene expression upon transfection into mammalian cells and suggests that altering the digestion or purification conditions to generate longer products is unnecessary for the efficient reduction of target gene expression.
Figure 2. 12-15 bp RNase III Digestion Products Elicit Silencing. A 200 bp GAPDH dsRNA (30 g) was digested with RNase III (30 U) for 1 hour at RT. Digestion products were run on a 15% non-denaturing acrylamide gel and the 12-15 bp products were excised, eluted, and ethanol precipitated. A sample was run on a 15% non-denaturing acrylamide gel for visualization (2A). HeLa cells were transfected with 100 nM of the 12-15 bp RNase III generated GAPDH siRNAs or a 21 bp chemically synthesized GAPDH siRNA. GAPDH protein levels were monitored by immunofluorescence 48 hours after transfection (2B) and the resulting images were quantitated (2C).
Specificity of Gene Silencing. We next analyzed the specificity of the siRNA for reducing target gene expression. HeLa cells were transfected with an RNase III generated siRNA population to GAPDH, and the resulting expression levels of GAPDH and a number of nonspecific target genes (La, Ku-70, c-myc, -actin, and cdk-2) were compared in transfected and nontransfected cells. Figure 3 shows a 63% reduction in GAPDH levels but no detectable reduction in the other genes examined. These data s uggest that nonspecific gene silencing is not occurring in cells after transfection with RNase III generated siRNA cocktails. A recent article that examined the effect of RNase III generated siRNA cocktails on related RNA binding proteins confirms the lack of nonspecific effects (5). In non-mammalian systems, mixtures of siRNAs to multiple sites within the target gene are generated in vitro by Dicer digestion of long dsRNA, which in turn specifically silence target genes. This suggests that there are mechanisms in place to maintain a high degree of specificity.
Comparison of RNase III Generated siRNAs to Individual Chemically Synthesized siRNAs. To compare the gene silencing effects of siRNA cocktails generated by RNase III versus individual chemically synthesized siRNAs, we transfected HeLa cells with siRNAs targeting GAPDH generated by both methods at 50 nM, 25 nM and 12.5 nM final concentration. The resulting protein levels were examined 48 hours after transfection. siRNAs prepared by both methods efficiently reduced GAPDH protein levels in a dose dependent manner, although higher concentrations of RNase III-generated siRNAs were required to maximally reduce GAPDH expression levels (data not shown). This finding is not surprising given the fact that siRNA cocktails generated by RNase III consist of both effective and ineffective siRNAs.
RNase III digestion of dsRNA is a robust, fast, and cost effective method of generating a population of siRNAs that can effectively silence target genes. The siRNA population generated by this method does not exhibit higher levels of toxicity (data not shown) or nonspecific effects on gene expression as compared to individual chemically synthesized siRNAs. The use of RNase III to prepare siRNA cocktails eliminates the need for expensive chemical synthesis, laborious siRNA expression vector production, or the time-consuming screening necessary to identify an effective siRNA sequence. RNase III generated siRNA cocktails are a good alternative to standard methods of siRNA production for the researcher that needs a quick result and does not require siRNA sequence information.
To simplify the preparation of siRNAs by RNase III digestion, Ambion now provides the Silencer siRNA Cocktail Kit (RNase III). This kit includes reagents and detailed protocols for generating and purifying dsRNA, dsRNA digestion by RNase III, and subsequent siRNA clean up. For more information, see the article Avoid siRNA Design Altogether.
Figure 3. RNase III siRNA Cocktails Show Specificity for Silencing. HeLa cells were transfected with 100 nM RNase III generated siRNAs to GAPDH. Immunofluorescence analysis of GAPDH, La, c-MYC, Cdk-2, Ku-90, and -actin was performed 48 hours post transfection and subsequently quantitated.
Materials & Methods
Preparation of siRNA Cocktails with RNase III Total RN A was extracted from HeLa cells (RNAqueous Kit, Ambion) and reverse transcribed to produce cDNA (RETROscript Kit, Ambion). PCR primers containing T7 RNA polymerase promoters were designed to amplify a 200 bp fragment approximately 200 bp from the 5' end of each gene of interest: human GAPDH, La, and c-fos. After PCR, the resulting templates were used in the Silencer siRNA Cocktail Kit (RNase III) to prepare siRNA cocktails to the individual genes according to the kit protocol. Briefly, the templates were used in an in vitro transcription reaction to generate dsRNA. After a brief column purification step, 15 g of the resulting dsRNA was digested with 15 U of RNase III at 37C for 1 hour. The digestion products were then purified with the siRNA Purification Units included in the kit to remove any undigested dsRNA. The resulting siRNA population was quantitated using a spectrophotometer and visualized on a 20% non-denaturing acrylamide gel.
Transfections HeLa cells at 30,000 cells per well, or 293 cells at 50,000 cells per well, were grown on glass coverslips in a 24 well tissue culture plate and transfected with siRNA at the indicated concentrations using siPORT Lipid (Ambion).
Immunofluorescence Analysis Immunofluorescence was performed on each sample after 48 hours, using specific primary antibodies (anti-GAPDH from Ambion; anti-La from Transduction Laboratories; anti-c-FOS from Santa Cruz Biotech). A FITC-conjugated donkey anti-mouse IgG secondary antibody (Jackson Immuno Research) was used for all experiments. All samples were mounted on slides using VectaShield with DAPI (Vector Laboratories) to allow for visualization of the cellular nuclei, and the resulting fluorescence microscopy images were digitally captured and quantified using Metamorph software (Universal Imaging Corp.).
Size Separation of RNase III Products After a 15 minute digestion at room temperature, reaction products were separated on a 15% non-denaturing acrylamide gel. 12-15 bp region was excised and eluted in Probe Elution Buffer (Ambion) for 18 hr at 37C, ethanol precipitated and resuspended in nuclease free water.