|Home | About | Journals | Submit | Contact Us | Français|
Chemically synthesized small interfering RNAs (siRNAs) are tools used for silencing the expression of a single gene. They are mainly employed in basic research applications, but may also have great potential in therapeutic applications. Longer double-stranded RNAs, such as Dicer-substrate 27mers, trigger gene silencing through the intrinsic RNAi pathway. The design of these Dicer-substrate 27mers has been optimized so they can be oriented by Dicer to consistently select the antisense (guide) strand after cleavage to shorter siRNAs, leading to predictable mRNA cleavage. In this paper we describe evidence that these Dicer-substrate 27mers produce more potent and sustained gene silencing for four genes when compared with synthetic 21mers that have the same guide-strand sequence. Furthermore, improved silencing by these 27mers is often more pronounced at lower concentrations.
Chemically synthesized small interfering RNAs (siRNAs) are used as tools for silencing the expression of a single gene by triggering the cellular RNA interference (RNAi) pathway.1–3 These synthetic siRNAs are widely used for basic research applications, but also have potential in therapeutic applications.4,5 Traditionally, synthetic siRNAs are composed of double-stranded RNAs (dsRNAs) that are 19 to 22 base pairs (bp) long with two-base 3′ overhangs. These short dsRNAs do not serve as substrates for the RNase III enzyme Dicer.6 Introduction of longer dsRNAs, which can be processed by Dicer, into the cell also triggers gene silencing through the intrinsic RNAi pathway, which is conserved in nearly all biological systems.7–10
Models of the intrinsic RNAi pathway describe a sequence of molecular interactions that lead to gene silencing (Figure 1).11–13 The pathway is activated with the introduction of an RNAi-initiating molecule, a long dsRNA such as Dicer-substrate 27mers.14 This is followed by the formation of an initial complex containing dsRNA, Dicer, and TAR RNA binding protein (TRBP) (R2D2 in D. melanogaster).2,13 Dicer is a complex protein comprised of a dsRNA binding domain, a PAZ domain, an RNA helicase domain, and two RNase III class domains.2,15,16 Dicer binds and cleaves the long dsRNA in the cell, producing short 21–23 nt duplexes with 5′ monophosphates and two-base 3′ overhangs. 17–19 The orientation of Dicer binding helps to select the antisense (guide) strand of the siRNA.18 This guide strand is used to identify the target mRNA by the mature RNA-induced silencing complex (RISC).20 Proteins in the Argonaute family then cleave the target mRNA, reducing mRNA levels and potentially attenuating the associated protein’s expression, leading to gene knockdown.15,16,21,22
In an effort to improve silencing efficiency, Kim et al.16 assayed a variety of synthetic siRNAs and Dicer-substrate dsRNAs of various lengths for effective gene silencing in HEK293 cells. Of all the lengths tested, one Dicer-substrate 27mer that was blunt ended was found to be up to 100-fold more potent at silencing the target gene when compared to both shorter (21–23 bp) and longer (35–45 bp) dsRNAs. However, blunt-ended 27mers can create inconsistent products and results.6 To leverage the enhanced potency of the 27mers, while avoiding the unpredictable performance of the blunt ends, improvements to the 27mer design were investigated. Some of the design improvements discovered led to further increases in gene silencing and more consistent results that are likely due to the nature of interactions between Dicer and 27mers.18 The 27mer design changes included adding a single 3′ overhang on the antisense strand, and placing two DNA bases on the 3′ end of the sense strand to orient Dicer binding and cleavage,18,23 resulting in the predictable sequence of the processed siRNA and subsequent selection of the antisense strand (Figure 2). For the remainder of this article, we will refer to this newly optimized designed molecule as Dicer-substrate 27mer.
In contrast to Dicer-substrate 27mers, the traditional synthetic siRNAs mimic the product of Dicer cleavage. However, synthetic siRNAs can be less effective at gene silencing, possibly because they do not engage Dicer with the same affinity and are not passed from Dicer to RISC with strand-specific orientation (Figure 1B).6,18
In this paper, we present further evidence that Dicer-substrate 27mers produce more potent and sustained gene silencing when compared with 21mers that have the same functional guide-strand sequence in RISC. siLentMer validated Dicer-substrate siRNAs are 27mers that produce a reduction in gene expression of 85% or more in HeLa cells,24 increasing the likelihood of success in any gene-silencing experiment.
HeLa cells were cultured in DMEM (Invitrogen, Carlsbad, CA) that was supplemented with 10% FBS (Thermo Fisher Scientific, Waltham, MA), sodium pyruvate, and nonessential amino acids. Cell cultures were maintained between 50 and 90% confluency. For the dose-response experiments, HeLa cells (ATCC number CCL-2) were seeded at a concentration of 1.1 × 104 cells/well in 48-well microplates in 250 μL of media per well. Cells were harvested 24 h posttransfection.
For time-course experiments, HeLa cells were seeded at a concentration of 1.1 × 104 cells/well in 24-well microplates with 500 μL of media per well. At 2 d posttransfection, one set of 24-well plates was split 1:3 into two additional sets of 24-well plates with 500 μL of media per well for RNA extraction at 2 and 6 d posttransfection. Cells were then harvested at 2, 4, and 6 d post-transfection.
All chemically-synthesized dsRNA sequences for the 21mers and Dicer-substrate 27mers used in this study were synthesized and HPLC purified (Integrated DNA Technologies, Coralville, IA; validated 27mers are offered by Bio-Rad Laboratories, Hercules, CA). The dsRNA sequences used for the dosage response and silencing longevity experiments included 21mers and Dicer-substrate 27mers with the same guide-strand sequence (Table 1). Non-silencing dsRNA controls (NC) included 27mers (Integrated DNA Technologies) and 21mers (Thermo Fisher Scientific).
Transfections were performed in 48- or 24-well plates using siLentFect lipid reagent (Bio-Rad Laboratories) for delivery of Dicer-substrate 27mers and 21mers into HeLa cells. Prior to application, the lipid:dsRNAs mix was diluted in OptiMEM-reduced serum media (Invitrogen) and allowed to incubate at room temperature for 40 min.
For the dose-response experiments, the final concentrations of 21mers and validated Dicer-substrate 27mers applied to the cells was equal to 0.0, 0.1, 1.0, 5.0, and 50 nM. The final lipid concentration was 0.2 μL in 275 μL of media.
For the time-course experiments, the final concentration of 21mers and validated Dicer-substrate 27mers was 5 nM and the lipid concentration was 0.5 μL in 550 μL of media.
RNA was extracted using the ZR-96 Mini RNA Extraction kit (Zymo Research, Orange, CA) according to the instructions, with one modification. After the first RNA wash step, a one-column DNA digestion was performed using DNase I, DNase I dilution buffer, and high-stringency wash buffer from the Aurum RNA Mini kit (Bio-Rad Laboratories). RNA concentration was measured using the Nanodrop ND-1000 spectrophotometer (Thermo Fisher Scientific). cDNA was synthesized with the iScript cDNA Synthesis kit (Bio-Rad Laboratories) according to the instructions, using 150 ng total RNA per sample. After cDNA synthesis, 20 μL of the cDNA synthesis reaction volume was diluted in 100 μL of water.
After treatment of cells with 21mers and Dicer-substrate 27mers, the PCR primers listed in Table 2 were used to assay gene expression by quantitative PCR (qPCR).
Real-time PCR reactions and analysis were run using the iQ5 Multicolor Real-Time Detection system (Bio-Rad Laboratories). Real-time PCR reactions included 12.5 μL of 2X iQ SYBR Green Supermix (Bio-Rad Laboratories), 10.5 μL of diluted cDNA, and 2 μL of a mixture containing the gene-specific forward and reverse primers (Table 2). All real-time PCR reactions were performed in quadruplicate using the three-step PCR program shown in Table 3.
Three independent cell-culture experiments were used to derive the average percentage of silencing by real-time PCR. Real-time PCR data was analyzed using iQ5 software (Bio-Rad Laboratories) and Excel spreadsheets (Microsoft, Redmond, WA).
HeLa cells were transfected with 21mers and validated Dicer-substrate 27mers that contained the same guide-strand sequence. The potency of silencing was determined by calculating the percent of remaining mRNA expression in cells treated with Dicer-substrate 27mers, 21mers, or the appropriate non-silencing negative controls (NC). The cells were assayed for potency of silencing in a dose-response experiment (Figure 3) and duration of action in a silencing-longevity experiment (Figure 4). The results of both experiments showed improved gene silencing (Figures 3A and and4A)4A) or similar gene silencing (Figures 3B and and4B)4B) when using Dicer-substrate 27mers for the five target genes listed in Table 1.
The dose response of HeLa cells to siRNAs was assayed using RT-qPCR after transfection with a range of 21mers and validated Dicer-substrate 27mers at four concentrations (100 pM, 1 nM, 5 nM, and 50 nM). Validated Dicer-substrate 27mers produced significantly better silencing for four of the five target genes (Figure 3A) over the range of dsRNA concentrations. Validated Dicer-substrate 27mers were also more effective at silencing target genes at very low concentrations (100 pM, 1 nM, and 5 nM), offering the advantage of using less dsRNA per experiment.
The longevity of silencing was examined by assaying gene expression using RT-qPCR for up to 6 d after treatment with 5 nM of 21mers or validated Dicer-substrate 27mers (Figure 4). When compared with 21mers, Dicer-substrate 27mers showed substantially longer silencing for four of the five target genes tested (Figure 4A), offering the advantage of longer silencing at a lower dosage. For the AKT1 gene, Dicer-substrate 27mers reduced gene expression to a level similar to that of 21mers (Figure 4B).
This paper compares gene silencing by Dicer-substrate 27mers to synthetic 21mers for five genes. The longer 27mer duplexes are cleaved by Dicer in vivo to yield a product with a guide-strand sequence that is identical to the synthetic 21mers. These two types of RNAi initiating molecules were transfected into HeLa cells in parallel to test for potency and longevity of silencing.
The dose-response experiments show a greater reduction in transcript level in four of the five genes tested when using a Dicer-substrate 27mer (Figure 3A). One gene, AKT1, showed similar levels of gene expression for both RNAi initiating molecules (Figure 3B). Interestingly, the improved gene silencing when using Dicer-substrate 27mers was often pronounced at the lower concentrations (Figure 3A; CDK2, TP53, and ACTB). For example, TP53 displayed a dramatic difference in silencing between the 21mers and 27mers at 1 nM concentration, while at 5 nM the difference in silencing was far less pronounced. This would indicate that there is a real advantage in using Dicer-substrate 27mers when compared to their matched synthetic 21mers because they require a lower concentration to achieve the same level of silencing. Low dosage is also important to avoid off-target effects,6,25–28 such as spurious gene silencing and interferon induction. For example, these adverse effects are known to be reduced by lowering the concentration of synthetic 21mers. The increased potency gained by using Dicer-substrate 27mers makes them an attractive option for those interested in developing RNAi-based therapies.
In a related dose-response result, the level of silencing by Dicer-substrate 27mers did not substantially improve when the 27mer concentration was increased from 5 to 50 nM in all genes (Figure 3). This observation can be explained by either of two models. First, the method of transfection (lipid mediated) was not capable of increased delivery above the 5 nM concentration. This is not likely to explain the result because the level of silencing for synthetic 21mers transfected by the same method continued to show improved silencing for four of the five genes (CDK2, TP53, RAF1, and ACTB). Second, for every gene there is a maximum possible level of silencing based on (1) the speed at which transcripts can be degraded, compared to the rate of transcription, (2) shielding of some portion of the transcript pool from cleavage or degradation, or (3) a combination of both. This area of research will be the focus of future studies.
Dicer-substrate 27mers also displayed advantages over synthetic 21mers when gene silencing was measured over time (Figure 4). The 27mers maintained high levels of gene silencing in four of the five genes when tested out to 6 d post-transfection. The matched synthetic 21mers showed a dramatic loss of silencing by 4 d post-transfection in TP53 and CDK2, while ACTB and RAF1 clearly showed moderate loss of silencing at 6 d post-transfection. Even AKT1 displayed marginally better gene silencing at 4 and 6 d post-transfection when using a 27mer compared to a 21mer (Figure 4B), although there was no difference between 27mers and 21mers silencing in the dose-response experiment (Figure 3B). Additionally, the shortest optimum time point for gene silencing was 2 d for either Dicer-substrate 27mer or 21mer RNA initiating molecules.
There are many unpredictable variables in an RNAi experiment, and successful gene silencing is heavily dependent on attributes of the dsRNA sequence, the gene sequence, and properties of that gene’s regulation. Given the advantages of the Dicer-substrate 27mers and the availability of validated products that are pre-tested for silencing performance, these RNAi initiating molecules have great potential to contribute to further advancements in RNAi research. Therefore, using a validated Dicer-substrate 27mer will eliminate the need to search for sequences that effectively silence specific genes. For example, siLentMer siRNAs are validated through a rigorous process that includes quantitating gene silencing in HeLa cells to verify silencing that is greater than or equal to 85%.24 During this validation process, several potential dsRNA sequences are tested for performance to ensure a more predictable silencing result in subsequent experiments. This predictability may help offset the capricious nature of some gene expression analyses.