Similar to C lividipennis RDL Jiang et
Similar to C. lividipennis RDL (Jiang et al., 2015), we found that the intracellular loop variants of A. mellifera RDL had no effect on GABA EC50 (Table 1). Also, as with C. lividipennis, there was no significant difference in maximal GABA responses (Table 1). However, whereas the 31 amino acid insertion in C. lividipennis RDL resulted in a decrease in fipronil sensitivity (Jiang et al., 2015), we observed that the three honey bee RDL variants had similar fipronil IC50 values (Table 1). We therefore speculate that the insertion sequence, as opposed to disruption of the putative TVR phosphorylation site, had an effect on insecticide sensitivity. These findings highlight an intriguing species-specific route to altering target site sensitivity to insecticides. Thus, even though the site of insertion is highly conserved, the actual sequence introduced may mediate certain effects dependant on the insect species. It will be of interest to see whether RDL of pest species also show a similar mechanism to diversify the Rdl transcriptome. If so, these variants can be monitored to determine whether the expression of certain isoforms is altered in association with insecticide exposure/resistance, thereby identifying novel routes to insecticide insensitivity in the field. Furthermore, perhaps these species-specific variants can be exploited to develop strategies to control pests whilst sparing non-target organisms. For example, sequences of these variants in pest species could be targeted by RNA interference, a newly emerging technology for crop protection (Lombardo et al., 2016). It remains to be determined what the roles are for the RDL intracellular loop variants. Other insect cysLGICs also have intracellular loop variants, such as the long and short variants of the A. mellifera α3 nicotinic NHS Biotin australia receptor subunit (Jones et al., 2006), the function of which is also unknown. In this regard, it is interesting to note that the vertebrate GABAA γ2 subunit has short (γ2S) and long (γ2L) variants differing by an 8 amino acid insertion in the intracellular loop, the long version of which has a potential protein kinase C phosphorylation site (Whiting et al., 1990). γ2S and γ2L show different distribution patterns in the adult brain (Gutierrez et al., 1994). Green fluorescent protein-tagging studies revealed that γ2L displayed a significantly higher capacity to accumulate at inhibitory synapses than γ2S, with phosphorylation of the protein kinase C site in γ2L playing a role in this postsynaptic clustering (Meier and Grantyn, 2004). Work on cell lines indicate that γ2S can act as an accessory protein, modulating functional properties of GABAA receptors without being incorporated in the ion channel-forming pentamer (Boileau et al., 2010). Several studies have shown that specific sequence or structural motifs in the intracellular loop mediate interaction with other proteins. For example, the 14-3-3 adaptor protein appears to bind the α3 nAChR loop at an RSSSSES consensus sequence or interacts with a coiled-coil motif (Rosenberg et al., 2008). Concerning RDL, further experiments are required to confirm whether the intracellular loop and its variants play similar important roles in insects. Also, future studies are required to determine which sites in the intracellular loop of insect cysLGICS are phosphorylated. We show that the neonicotinoid, imidacloprid, acts directly on A. mellifera RDL (Fig. 5) as an antagonist, consistent with findings that it reduced GABA-induced responses in cultured honey bee Kenyon cells (Deglise et al., 2002). All three Amel_RDL variants showed similar sensitivity to imidacloprid. We did, however, observe that the honey bee RDL was less sensitive to the antagonistic effects of imidacloprid than An. gambiae RDL (Fig. 5B). The A296G mutation in TM2 of An. gambiae RDL abolished imidacloprid action, suggesting the neonicotinoid is acting as a channel blocker (Taylor-Wells et al., 2015). The TM2 peptide sequences of A. mellifera and An. gambiae RDL are identical, indicating that residues other than those lining the ion channel are involved in the differential sensitivities of the two species. Since neonicotinoids interact with the agonist binding domain in nAChRs (Ihara et al., 2015), perhaps the action of imidacloprid on RDL is influenced by residues in the N-terminal extracellular domain. The concentration of imidacloprid required to antagonise A. mellifera or An. gambiae RDL is notably high and it remains to be determined whether RDL plays any role in the insecticidal effects of neonicotinoids. Novel and selective chemistry that both activates nAChRs and inhibits GABA receptors, or a combination of molecules with these activities, could be additive in the impact on neuronal excitability and represent a new generation of improved insecticides.