Not two, but three mutations. The first & second induced mutations gives the fruit fly immunity to the poison but the mutations are also lethal, which means the fly could not survive long enough for the third needed mutation to occur in an adaptive walk. The third mutation is like a stabilizer of the first & second mutation, it does not give immunity. In the test all three mutations are needed simultaneously for the fly to have the immunity and survive, that is not an adaptive walk, that is incredible engineering.
So essentially you can engineer a monarch butterfly, but an adaptive walk is impossible in this case. (Writing on the go, may edit later)
ceejayoz|6 years ago
> In multiple lineages, the substitutions A119S and A119N preceded substitutions to 111 and 122 (Hemiptera, Hymenoptera, Diptera). In Drosophila, where we have the greatest phylogenetic resolution, A119S was established before substitutions to sites 111 and 122 in the evolutionary lineage leading to D. subobscura, which appears to be polymorphic with respect to CG-insensitivity.