Detrimental effects of RNAi: a cautionary note on its use in Drosophila ageing studies

Abstract

RNA interference (RNAi) provides an important tool for gene function discovery. It has been widely exploited in Caenorhabditis elegans ageing research because it does not appear to have any non-specific effects on ageing-related traits in that model organism. We show here that ubiquitous, adult-onset activation of the RNAi machinery, achieved by expressing a double stranded RNA targeting GFP or lacZ for degradation, or by increasing expression of Dicer substantially reduces lifespan in Drosophila melanogaster. Induction of GFPRNAi construct also alters the response of lifespan to nutrition, exacerbating the lifespan-shortening effects of food containing a high quantity of yeast. Our study indicates that activation of the RNAi machinery may have sequence-independent side-effects on lifespan, and that caution needs to be exercised when employing ubiquitous RNAi in Drosophila ageing studies. However, we also show that RNAi restricted to certain tissues may not be detrimental to lifespan.

Figure 1. Ubiquitous adult-induced expression of dsRNA targeting GFP for RNAi shortens lifespan in Drosophila females.

Mated female flies of the indicated genotype were fed 1 SYA food containing RU486 (red lines) or not (black lines) from day two of adulthood. A and B Log-rank test detected significant differences between the induced and uninduced conditions (p<0.0001, n =  ∼100 per condition). C, D and E Log-rank test did not detected significant differences between the induced and uninduced conditions (p>0.05, n =  ∼100 per condition).

Figure 2. Ubiquitous adult-induced expression of dsRNA targeting lacZ for RNAi, or of Dicer2, shortens lifespan in Drosophila females.

A Mated ActGS>lacZRNAi female flies of the indicated genotype were fed 1 SYA food containing RU486 (red lines) or not (black lines) from day two of adulthood. Log-rank test detected significant differences between the induced and uninduced conditions (p = 0.004, n =  ∼120 per condition). B Same as in A but for ActGS>Dcr2. Log-rank test detected significant differences between the induced and uninduced conditions (p = 0.0006, n =  ∼150 per condition).

Figure 3. Lifespan-shortening effects of ubiquitous adult-induced expression of UAS-GFPRNAi are diet-dependent.

Mated female tubGS>GFPRNAi flies were reared on 1 SYA food and placed on different foods containing increasing amounts of yeast (0.1× to 2× relative to standard food) either containing RU486 (shades of red) or not (shades of grey) from day two of adulthood. A The resulting lifespans. B The median survival times, with asterisks indicating significant differences in survival between the induced (+ RU486) and uninduced (-RU486) conditions (Log-rank test, p<0.001, n =  ∼150 per condition). The survival between induced and uninduced at 0.5× yeast was not significant (Log-rank test, p>0.5, n =  ∼150 per condition). Cox Proportional Hazards analysis revealed significant effect of yeast quantity, presence of RU486 and their interaction (p<0.0001).

Figure 4. Adult-induced expression of dsRNA targeting GFP for RNAi in midgut and abdominal fat body or in neurons does not shorten lifespan in Drosophila females.

Mated female flies of indicated genotype were fed 1 SYA food containing RU486 (red lines) or not (black lines) from day two of adulthood. A Log-rank test detected a slight but significant increase in lifespan in the induced compared to the uninduced condition (p = 0.04, n =  ∼80 per condition). B Log-rank test did not detected significant differences between the induced and uninduced conditions (p>0.05, n =  ∼80 per condition).