The SID-1 double-stranded RNA transporter is not selective for dsRNA length

Abstract

The double-stranded RNA (dsRNA) transport protein SID-1 enables systemic RNA interference (RNAi) in Caenorhabditis elegans, whereby silencing initiated by local exposure to dsRNA spreads throughout the animal and to its progeny. Previously, we showed that providing dsRNA in the growth medium of Drosophila S2 cells that express C. elegans SID-1 efficiently triggers RNAi. In these experiments long dsRNA proved to be significantly more effective than short dsRNA in silencing the target gene. Here, we show that equivalent masses of long or short dsRNA accumulate in these cells, indicating that size-dependent silencing is not due to size-selective transport through SID-1. Furthermore, using pulse-chase dsRNA uptake experiments, we show that short dsRNA accumulates more rapidly than long dsRNA. We found that import rates are dependent on dsRNA concentration, consistent with energy-independent, diffusion-limited transport through the SID-1 channel. Comparison of silencing efficiencies between Drosophila S2 cells heterologously expressing SID-1 and primary-cultured C. elegans cells shows similar dsRNA concentration and size dependencies, suggesting that C. elegans regulatory proteins do not measurably enhance or restrict dsRNA transport through SID-1. Finally, we find that coexpressing mutant SID-1 with wild-type SID-1 in S2 cells interferes with SID-1 function, indicating that SID-1 may function as a multimer.

FIGURE 1.

Size-independent transport and retention of dsRNA. (A) dsRNA size-dependent silencing in Drosophila S2 cells expressing SID-1 (data adapted from Feinberg and Hunter 2003). Cells cotransfected with luciferase and SID-1 expressing plasmids were exposed to different concentrations of dsRNA of varying lengths and the extent of silencing was determined. The gray region (top) reflects average luciferase levels at the end of the experiment in samples treated with cycloheximide as a control for protein stability. (B) 10 ng/mL of 500 bp, 100 bp, and 50 bp dsRNA, and siRNA (internally labeled 500 bp dsRNA digested with Dicer) were added to induced and uninduced MT/sid-1 cells for 30 min. The mass of internalized dsRNA was normalized to 500 bp dsRNA. Error bars = 1 standard deviation, n = 2.

FIGURE 2.

Size- and concentration-dependent dsRNA transport by SID-1. (A) one or 10 ng/mL of radiolabeled 500 bp dsRNA was added to MT/sid-1 cells at time 0 and 100X unlabeled dsRNA was added at the indicated time points. Internalized radiolabeled dsRNA is reported as molecules/cell. Transport of radiolabeled 500 bp dsRNA shows a rapid initial phase (inset), which appears to be linear and concentration dependent in the first 10 sec. Error bars = 1 standard deviation, n = 4. (B) Retention at 10 and 300 sec of 100 bp and 500 bp dsRNA added at equimolar concentrations. Approximately twice as much 100 bp dsRNA was retained at both time points. For data presentation, CPM in uninduced cells were subtracted from the counts measured in induced cells. Error bars = 1 standard deviation, n = 2.

FIGURE 3.

Comparison of size and concentration-dependent silencing in C. elegans and Drosophila cells. Size and concentration-dependent silencing triggered by addition of dsRNA to the growth medium of cultured C. elegans and Drosophila cells expressing wild-type or mutant SID-1. (A,B) Cultured C. elegans cells expressing myo-3∷luciferase∷GFP were exposed to the indicated concentration of 500 bp (A) and indicated lengths (B) of luciferase dsRNA and luciferase levels measured 48 h later. The qt9 allele is an early stop codon mutant (Winston et al. 2002). (C,D) Drosophila S2 cells transiently transfected with wild-type SID-1 or missense mutant SID-1(S536I) constructs were exposed to the indicated concentration of 500 bp (C) and indicated lengths (D) of luciferase dsRNA and luciferase levels measured 24 h later. Error bars = 1 standard deviation, n = 3.

FIGURE 4.

Coexpression of SID-1(S536I) with wild-type SID-1 interferes with soaking RNAi. (A) Cotransfection of SID-1(S536I) with wild-type SID-1 reduces luciferase silencing. S2 cells were cotransfected with a constant amount of firefly luciferase plasmid and the indicated ratios of plasmids that express wild-type SID-1 and SID-1(S536I). A range of concentrations of 500 bp firefly luciferase dsRNA was then added to the growth medium. (B) SID-1(S536I) does not sequester dsRNA from wild-type SID-1 when expressed in different cells. S2 cells coexpressing firefly luciferase and wild-type SID-1 were cocultured at the indicated ratios with S2 cells expressing SID-1(S536I) with a range of concentrations of 500 bp firefly luciferase dsRNA added to the growth medium. For cells expressing SID-1(S536I) alone (0:1 ratio), firefly luciferase plasmid was cotransfected with the SID-1(S536I) expressing plasmid. Error bars = 1 standard deviation, n = 3.