BACKGROUND AND PURPOSE Fluralaner is a novel isoxazoline insecticide with broad insect spectrum, and mainly acts on the insect GABA receptor with unique binding action, but its molecular interaction with insect GABA receptor has not been deeply identified on molecular level according to its selectivity between target (insect) and non-target (mammal) organisms. EXPERIMENTAL APPOACH The potential binding residues (I258T and L275I in TM1; V288I, M298N, G303N and A304S in TM2; G3’M/S, A327S, G336N, M338I and A339F in TM3; M473V and I477D in TM4) were predicted by SYBYL-X 2.1 software, and verified respectively by the site-directed mutagenesis and two-electrode voltage clamp (TEVC) technique. KEY RESULTS In the 11 predicted amino acids, the G3’M has the strongest ability to reduce the sensitivity of recombinant rice stem borer RDL homomeric channel to fluralaner. Compared with the wild-type (WT)-RDL, the G3’M mutation almost completely abolish the binding of fluralaner and avermectin, but not fipronil on recombinant homomeric channel of RDL from several orders of insects in vitro. In addition, the M3’G on rat Mus musculus β2 improved the sensitivity of recombinant heteromeric Mmα1β2-M3’G channel to fluralaner. Our results demonstrated that the glycine at the third position of TM3 determines the action of fluralaner and should be the binding site of fluralaner with RDL. CONCLUSION AND IMPLICATIONS These results would contribute to understanding the molecular interaction of fluralaner with RDL homomeric channel and may be used to guide future modification of isoxazolines to achieve highly selective control of pests with minimal effects on non-targeted organisms.

Target-site insensitive mutations and overexpression of detoxification genes are two major mechanisms conferring insecticide resistance. Many molecular assays were applied to detect these two kinds of resistance genetic markers in insect populations. Unfortunately, these assays are time-consuming and have high false-positive rates. RNA-Seq data, which contains information on the variation within expressed regions of the genome and expression information of detoxification genes, provides us a valuable resource to detect resistance-associated markers. At present, there is no corresponding method at present. Here, we collected 66 reported resistance mutations of four main insecticide targets (AChE, VGSC, RyR, and nAChR) of 82 insect species. Next, we obtained 403 sequences of the four target genes and 12,665 sequences of three kinds of detoxification genes including P450, GST, and CCE. Here, we developed a Perl program, FastD, to detect insecticide target-site insensitive mutations and overexpressed detoxification genes from RNA-Seq data, and constructed a web server for FastD (http://www.insect-genome.com/fastd). FastD program was then applied to detect two kinds of resistant markers in five populations of two insects, Plutella xylostella and Aphis gossypii. Results showed that RyR mutation G4946E was detected in all P. xylostella populations, with higher frequencies in two resistant populations, ZZ (66.1%) and CHR (94.55%), than a susceptible population CHS (2.32%). CYP6a2 was overexpressed 10.82-fold in ZZ population. As to A. gossypii, nAChR mutation R81T was detected in resistant population KR with 49.85% frequency, but not in susceptible population NS. CYP6CY22 and CYP6CY13 were overexpressed 39.61- and 22.04-fold respectively in KR population. FastD is a program using RNA-Seq data to detect two types of resistance markers to estimate resistance level of insect populations. Generally, resistance level estimated by FastD were consistent with previous reports, confirming the reliability of this program in predicting population resistance at omics-level.