The diamondback moth, Plutella xylostella, is a serious agricultural pest of Brassica crops worldwide and can be extremely difficult to control. Populations commonly evolve resistance to insecticides used against them, including biological based Bt toxins produced by the bacteria Bacillus thuringiensis. Hundreds of different Bt toxins have been described, and many are highly specific to targeted pests yet benign to non-target organisms¹ . For example, Cry1Ac kills lepidopteran moth pests, but not flies (e.g. Drosophila), or beneficial insects such as parasitoids.

The precise mode of action of Bt toxins remains controversial² . Two well supported models, the ‘Classic Model’ and ‘Sequential Binding Model’, both result in midgut pore formation yet differ in the types of toxin receptors required. Although many Bt midgut receptors have been proposed in the literature (alkaline phosphatase³ , aminopeptidases⁴ , cadherin⁵ ), mutations in the membrane bound ABC transporter C2, have recently been associated with resistance to Bt Cry1A toxins in multiple Lepidoptera⁶⁷⁸.

As Drosophila is not susceptible to Cry1Ac, we used this model system to investigate whether the diamondback moth ABCC2 protein acts as a functional toxin receptor. The moth ABCC2 gene was fused with a GFP reporter gene, cloned into the pUAST vector, and transgenic Drosophila lines then generated. Using the GAL4/UAS system, we observed successful expression of the ABCC2-GFP construct, through GFP localization to cell membranes. When fed with artificial diet containing Cry1Ac toxin, larvae expressing ABCC2 in the midgut showed 100% mortality, while expression in salivary glands had no affect on survival. We aim to further use this system to help understand Bt toxin mode of action, through adding support to either the Classic Model or Sequential Binding Model.