The advent of RNA interference (RNAi) has facilitated a boom in biochemical pathway analysis and functional genomics. RNAi refers to any RNA molecule that interferes with the expression of its homologous gene product (1,2,3,4,5). Also referred to as posttranscriptional gene silencing, RNAi is exquisitely specific for the targeted gene and encompasses sense, antisense, or double-stranded RNA (dsRNA) molecules, although it is commonly attributed with dsRNA as single-stranded RNA (ssRNA) effects have been traced to low levels of contaminating dsRNA (3,4).

dsRNA of 220–700 bp has been shown to significantly reduce levels of messenger RNA (mRNA) transcript in Drosophila cell culture for a wide range of genes including insulin signaling pathway components (6), recombinant green fluorescent protein (GFP) (7), and for 91% of the genes associated with proliferation and survival (8). dsRNA of 500–700 bp can be transfected into Drosophila Schneider 2 (S2) cells by incubating them with fetal bovine serum (FBS) following serum starvation (6). However, FBS has numerous stimulatory effects and can greatly complicate metabolic studies, due its poorly characterized and variant composition. Recent work has taken advantage of serum-free media (SFM) for metabolic studies in cell culture (9,10). Unfortunately, while S2 cells can grow without serum, we were unable to stimulate dsRNA uptake utilizing only serum-free media (data not shown).

In this study, we were able to remove FBS completely from dsRNA transfection by replacing it with bovine insulin. To test the effect of trying to replace FBS with insulin, we selected Cyclin E as a target gene for silencing. To test the effect of varying the amount of dsRNA used for silencing, we selected the tuberous sclerosis complex gene (TSC1) as a target. S2 cells were grown in Drosophila SFM (Invitrogen, Carlsbad, CA, USA). dsRNA was synthesized following a modified version of the method developed by Clemens and coworkers (6). Briefly, S2 cells were grown to 5 × 10⁶ cells/mL, and RNA was extracted using an RNAqueous® kit (Ambion, Houston, TX, USA) according to the manufacturer's protocol, including incubation with DNase to remove any contaminatng DNA. First-strand templates of Cyclin E DNA were synthesized from the total mRNA using oligo(dT) primers and a RETROscript™ kit (Ambion) as per the manufacturer's instructions. Oligo(dT) primers were selected to further minimize any effects from contaminating DNA. A 630-bp region of first-strand DNA template was PCR-amplified using Cyclin E-specific primers (5′-ATGGGTTTAAATGCCAAGAGTGTTTGTTC-3′; 3′-CACCACCACTGGCGTCTGCTTGCTTCCACG-5′) or a 700-bp region with TSC1-specific primers (5′-ATGACGCTGGAGA-ACGAGGAGGCCAAGCGC-3′; 3′-CCATCTCCTTCCATCGCG-TATTGTTTACC-5′). T7 sequences (5′-TAATACGACTCACTATAGGGA-3′) were added to each template using PCR, making T7 templates. To transcribe ssRNA, T7 templates were used with the MEGAscript™ kit (Ambion) as per the manufacturer's instructions. The ssRNA synthesized was extracted using phenol/chloroform and resuspended in nuclease-free water to a concentration of approximately 3.3 µg/µL. ssRNA was incubated at 65°C for 30 min before being allowed to cool to room temperature on the benchtop. Subsequent to this annealing step, dsRNA was checked for size and integrity using agarose gel electrophoresis.

Cells were seeded in triplicate to 1 × 10⁶ cells/well in 0.7 mL SFM in 12-well plates and incubated with 15 µg/mL dsRNA against cyclin E, 15 µg/mL dsRNA against chloramphenicol acetyl transferase (CAT) (as a nonspecific control), or an equal volume of nuclease-free water. To determine the minimum amount of dsRNA needed for gene silencing and to demonstrate the generality of insulin-mediated uptake, cells were again seeded in triplicate to 1 × 10⁶ cells/well in 0.7 mL SFM in 12-well plates and incubated with 6, 15, or 30 µg/mL dsRNA against TSC1 or an equal volume of nuclease-free water. dsRNA (15 µg/mL) against DsRed served as a nonspecific control. CAT and DsRed were used as controls because they have no significant similarity to any genes in the Drosophila genome [Basic Local Alignment Search Tool (BLAST) search]. The CAT and DsRed dsRNA fragments were 630 and 689 bp long, respectively. After 1 h incubation, SFM containing either 10% FBS (Sigma, St. Louis, MO, USA) or 1.75 µM bovine insulin (Sigma) was added to a final volume of 2.1 mL/well. The amount of insulin used was determined to be the optimum for growth in S2 cell culture with SFM. Higher amounts were tried, but resulted in cell lysis as determined by trypan blue dye staining (data not shown). The plates were incubated for 60 h at 27°C. RNA was extracted from the cells, and reverse transcription PCR (RT-PCR) was performed as described above for a 630-bp segment of Cyclin E or a 700-bp segment of TSC1 and an 866-bp segment of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a loading control. Each treatment's triplicate RNA samples were pooled before reverse transcription. Results ((Figure 1)) indicate that insulin worked as well or better than FBS at facilitating posttranscriptional gene silencing in S2 cells and that the amount dsRNA needed for silencing using insulin has to be determined for each gene under study.

Figure 1.