Species-specific dsEGFP-derived small RNA profiles reveal divergence from computational siRNA predictions in insects:...

RNA interference (RNAi) efficiency varies widely among insect taxa, but whether this variation is associated with differences in the processing of double-stranded RNA (dsRNA) into small interfering RNA (siRNA) remains unclear. Here, we compared small RNA profiles generated from the same enhanced green fluorescent protein dsRNA (dsEGFP) in the highly RNAi-responsive red flour beetle Tribolium castaneum and the highly RNAi-refractory diamondback moth Plutella xylostella. Small RNA sequencing identified dsEGFP-derived 21-nucleotide (nt) siRNAs as the predominant size class among dsEGFP-mapped 21-25 nt small RNAs following incubation with dissected tissues from each species. Species-specific differences were evident in antisense siRNA abundance, positional distribution, and 5'-terminal nucleotide composition. P. xylostella generated fewer dsEGFP-derived antisense siRNAs than T. castaneum and the two species exhibited distinct antisense siRNA hotspots and 5'-terminal nucleotide preference. Comparisons with computationally predicted siRNAs from Designer of Small Interfering RNA (DSIR) and siDirect showed no exact sequence matches among the most abundant experimentally identified siRNAs, although positional overlap was evident, particularly with DSIR predictions. These findings demonstrate that dsRNA processing generates distinct small RNA profiles in insects with contrasting RNAi efficiencies. The low abundance of antisense siRNAs in P. xylostella suggests that species-specific processing of dsRNA may influence the pool of siRNAs available for RNA-induced silencing complex (RISC)-mediated loading and thereby contribute to RNAi efficiency. Thus, our results highlight the importance of empirical small RNA profiling as a complementary approach to computational prediction for identifying species-specific siRNA candidates and informing subsequent siRNA selection and functional validation in insects.