Background: L. decemlineata is an exotic invasive insect pest, and invaded in Xinjiang Uygur autonomous region in China in the 1990s from Kazakhstan. It is a notorious defoliator of potato throughout most of the northern Xinjiang in current, and often causes extremely large yield losses of potato. Results: The expression stability of nine L. decemlineata house-keeping genes (Actin, ACT1 and ACT2; ADP-ribosylation factor, ARF1 and ARF4; TATA box binding protein, TBP1 and TBP2; ribosomal protein RP4 and RP18; translation elongation factor 1α EF1α) was evaluated by quantitative real-time polymerase chain reaction (qRT-PCR) in seven developmental stages, three larval tissues and two insecticide treatments. The results were analyzed using three software programs: geNorm, NormFinder and BestKeeper. Although there was no consistent ranking observed among the house-keeping genes across the samples, the overall analysis revealed that RP18, RP4, ARF1, and ARF4 were the four most stable house-keeping genes. In contrast, ACT1 and ACT2, two of the most widely used reference genes, had the least stability. Our results suggest that the combined use of the four most stably expressed genes may produce optimal normalization for qRT-PCR. Conclusions: The expression stability of the house-keeping genes varies among different developing stages, in different tissues and under different experimental conditions. Our results will enable a more accurate and reliable normalization of qRT-PCR data in L. decemlineata.

RNA interference (RNAi)-based technology has been proven as a novel approach for insect pest control. However, whether insects could evolve resistance to RNAi and the underlying mechanism is largely unknown. The target gene mutations were thought to be one of the potential ways to develop the resistance. Here we predicted the effective siRNA candidates that could be derived from dsRNA against the Colorado potato beetle (CPB) β-Actin gene (dsACT). By site-directed mutagenesis, we synthesized the dsRNAs with the defect in generation of effective siRNAs (and thus were supposed to have comparable low RNAi efficacy). We showed that, with mismatches to the target gene, all the dsRNA variants caused similar levels of silencing of target gene, mortality and larval growth retardation of CPB. Our results suggest that when the mismatch rate of dsACT and target β-Actin mRNA is less than 3%, the RNAi efficiency is not impaired in CPB, which might imply the low possibility of RNAi resistance evolving through the sequence mismatches between dsRNA and the target gene.