Introduction

Since the discovery of gene silencing in Caenorhabditis elegans (1), the importance of RNA molecules in the regulation of gene silencing has been rigorously studied (reviewed in Reference (2). The RNA interference (RNAi) process initiates with the cleavage of small interfering RNA (siRNA) precursors by Dicer followed by incorporation into the RNA induced silencing complex (RISC) (3,4,5). The production of siRNA duplexes leads to active RISC complexes whereby the bound guide (antisense) strand directs endonuclease cleavage of the mRNA by the argonaute protein (6,7,8). Thus, Dicer cleavage of dsRNA substrates—to yield small interfering siRNA molecules with a length of 21 bp containing a 3′2-nucleotide overhang—is an essential step in the RNAi process.

Gene silencing through the RNAi pathway can be initiated through the delivery of D-siRNAs followed by subsequent cleavage by Dicer or siRNA products. Recently, locked nucleic acid (LNA)–modified Dicer products have proven effective in vivo at producing gene knockdown (9,10). In the case of unmodified RNA, it has been demonstrated (11) that more efficient target mRNA cleavage can occur using Dicer substrates instead of the conventional 21-mer siRNAs. It has been hypothesized that the basis for the enhanced potency is the result of more efficient incorporation into RISC through the physical association of Dicer with the components of the of RNAi machinery (5,11). The identification of multiple protein cofactors known to interact with Dicer—and either functioning to suppress or enhance specific RNAi pathways—supports this proposal (12).

Increased interest in therapeutic applications for RNAi through the delivery of Dicer substrates brings about a need for a rapid method to screen large panels of potential D-siRNAs. Gel-based assays utilizing radiolabeled RNA are time-consuming and low-throughput. In order to circumvent these limitations, a continuous kinetic assay using a fluorescent Dicer probe substrate was developed. Unlike gel-based approaches for assessing Dicer cleavage, this fluorogenic Dicer assay allows for a rapid and quantitative assessment of potential Dicer substrates for their ability to be processed by Dicer in a purified enzyme system without the need for radiolabeling. The assay requires only a single fluorophore and quencher-labeled dsRNA probe to serve as a reporter for unlabeled test substrates. The structureactivity relationships obtained from this assay can be used to guide Dicer substrate design for potential RNAi-based therapeutics. In this report, a panel of 196 randomly selected D-siRNA substrates from human cDNA were screened using a fluorogenic Dicer substrate in heterologous competition assays to investigate the effect of 3′ overhang sequence in the cleavage efficiency of Dicer.

Materials and methods

RNA and expression/purification of protein constructs

Human Dicer was expressed in an insect baculovirus system using the titerless infected-cells preservation and scale-up method (13,14). Baculoviruses were generated using the Bac-to-Bac system (Invitrogen, Carlsbad, CA, USA). Dicer was engineered into pFastBac to express the full length protein with an N-terminal 6×His-tag, and a TEV protease site was inserted between the open reading frame and His-tag for purification purposes. Dicer was purified using a three-step purification method: nickel affinity, reverse nickel affinity, and size-exclusion chromatography. Briefly, 1 L cells was suspended in 100 mL lysis buffer (buffer NA) composed of 50 mM Tris pH 8.0, 500 mM NaCl, 5% glycerol, 0.1% Triton X-100, 5 mM TCEP, and protease inhibitors. The cells were lysed with a microfluidizer (Watts Fluidair, Richland, MI, USA). Cell debris was sedimented by centrifugation at 40,000× g for 2 h at 4°C . The supernatant was filtered and loaded with 5 mM imidazole onto a HisTrap HP column (GE Healthcare, Piscataway, NJ, USA) equilibrated with 0.5% buffer NB (buffer NA containing 1M imidazole). The column was step washed/eluted with 0.5% to 2 .5%, 2 .5% to 5–10%, and 5–10% to 30% buffer N B. The fractions containing Dicer were pooled and TEV protease was added (50 units/mg protein), followed by dialysis against Buffer NA overnight at 4°C to remove the His-tag. The dialyzed proteins were then applied to a HisTrap HP column (GE Healthcare) and the unbound proteins were pooled and concentrated to 3 mL . Samples were further purified through a size exclusion column Superdex200 16/60 (GE Healthcare Biosciences, Piscataway NJ, USA) equilibrated with gel filtration buffer (50 mM Tris pH 8.0, 300 mM NaCl, 0.5 mM EDTA , 10% glycerol, and 2 mM DTT) with a 1 mL/min flow rate. Fractions containing Dicer were pooled and concentrated. Protein concentrations were measured using the Bio-Rad Bradford assay (Cat. no. 500–0002; Hercules, CA, USA) with BSA as a standard. Proteins were flash frozen in liquid nitrogen and stored at −80°C . Labeled D-siRNA molecules were purchased from Integrated DNA Technologies (IDT, Coralville IA, USA) and purified by RNase-free high-pressure liquid chromatography (HPLC). All preparations of Dicer enzyme were tested for the presence of single-strand RNase activity with the RNase-Alert QC system (Cat. no. AM1966; Ambion, Austin, TX, USA) (Supplementary Figure 1). Gel electrophoresis analysis of Dicer cleavage products of 28-mer substrate confirmed a product size of ~21 bp.