The correct answer is (c) — the toxin is inactive.

Justification. The Bt toxin protein exists as inactive protoxins inside the bacterium. It is converted into the active form of toxin only once an insect ingests it, because the alkaline pH of the insect gut solubilises the crystals. Inside the bacterium there is no alkaline gut, so no activation, and therefore no harm.

Why each of the others is wrong. (a) The bacterium is not resistant — resistance would imply the active toxin is present but has no effect, which is not the case. (b) 'Immature' is the wrong word: the protein is complete, not half-built; it is a protoxin awaiting chemical activation, in the same way that a pro-enzyme is a complete molecule awaiting processing. (d) No special sac is involved; the toxin is present as protein crystals formed during a particular phase of growth, not sequestered in an organelle.

The wider significance. This same inactivity is why Bt crop material is safe to mammals, whose stomachs are acidic rather than alkaline.

What they are. Cry proteins are the toxic insecticidal proteins coded by the genes named cry. Bacillus thuringiensis forms protein crystals during a particular phase of its growth, and these crystals contain the insecticidal protein. Most Bt toxins are insect-group specific, so different cry proteins kill different groups: the proteins encoded by cryIAc and cryIIAb control the cotton bollworms, while that of cryIAb controls corn borer.

The organism. The bacterium Bacillus thuringiensis, Bt for short. Some of its strains produce proteins that kill certain insects such as lepidopterans (tobacco budworm, armyworm), coleopterans (beetles) and dipterans (flies, mosquitoes).

How man has exploited it. Specific Bt toxin genes were isolated from Bacillus thuringiensis and incorporated into several crop plants. The Bt toxin gene has been cloned from the bacteria and expressed in plants to provide resistance to insects without the need for insecticides — in effect creating a bio-pesticide. Examples are Bt cotton, Bt corn, rice, tomato, potato and soyabean. The choice of gene depends upon the crop and the targeted pest.

Why the plant, rather than a spray. The toxin is then produced only inside the plant tissue and only where the pest feeds, instead of being distributed over the field, the soil and the water draining from it — and it costs the farmer nothing per application, which matters most where agrochemicals are too expensive.

The strategy. Meloidegyne incognitia is a nematode that infects the roots of tobacco plants and causes a great reduction in yield. Using Agrobacterium vectors, nematode-specific genes were introduced into the host plant. The introduction of DNA was such that it produced both sense and anti-sense RNA in the host cells. Being complementary to each other, these two RNAs formed a double-stranded RNA that initiated RNA interference and silenced the specific mRNA of the nematode. The parasite could not survive in a transgenic host expressing specific interfering RNA, so the transgenic plant protected itself.

Why single-stranded RNA would not do. RNAi is triggered specifically by double-stranded RNA. The process involves silencing of a specific mRNA due to a complementary dsRNA molecule that binds to it and prevents translation. A single sense transcript is indistinguishable from ordinary cellular mRNA and initiates nothing; it would simply be translated or degraded. This is why the construct was deliberately designed to make both strands.

Why the cell recognises dsRNA as a danger signal. Because in nature it is one. RNAi takes place in all eukaryotic organisms as a method of cellular defence, and the natural sources of such complementary RNA are infection by viruses having RNA genomes, and transposons that replicate via an RNA intermediate. Double-stranded RNA in a eukaryotic cell normally means an invader.

How resistance arises. A Bt field exerts intense, continuous selection. Every bollworm that feeds there is exposed, and any rare individual whose midgut receptors bind the activated toxin poorly, or whose gut chemistry activates the protoxin less efficiently, survives and reproduces while the rest die. Over several generations, the frequency of such individuals rises. This is natural selection operating exactly as it does in the evolution of antibiotic resistance — the toxin does not create the resistant variant, it merely removes every competitor.

Why the neighbouring ordinary field matters. It is a refuge. Insects growing up there are never exposed to the toxin, so the susceptible form of the gene is preserved in the population at full frequency. When resistant survivors from your field mate with susceptible insects from the refuge, the offspring are usually not fully resistant, and resistance is diluted instead of accumulating. This is why planting a proportion of non-Bt crop alongside Bt crop is a deliberate resistance-management practice, not a lapse.

What else can be done. Because most Bt toxins are insect-group specific and there are a number of cry genes, a second toxin can be introduced into the same plant — as with cryIAc and cryIIAb, both of which control cotton bollworms. An insect would then need two independent rare changes at once, which is far less likely than one.

The conceptual point. A pest-resistance gene is not a permanent solution but a selection pressure. Any single-toxin strategy invites the evolution of resistance, and must be managed as such.

What the plant produces. In the Bt strategy the plant produces a foreign protein — an insecticidal Bt toxin, stored as inactive protoxin. In the RNAi strategy the plant produces RNA: both sense and anti-sense RNA, which pair to form double-stranded RNA. The plant itself is not made poisonous at all.

Specificity. Bt specificity operates at two levels — activation needs an alkaline gut, and the activated toxin must bind receptors on midgut epithelial cells — and most Bt toxins are insect-group specific, so the cry gene is chosen for the crop and the targeted pest. But the unit of specificity is the insect group: lepidopterans, coleopterans or dipterans, which may include harmless species. RNAi specificity is at the level of nucleotide sequence: only an mRNA complementary to the introduced dsRNA is silenced.

Which is less likely to affect non-target species. RNAi, in principle, because sequence complementarity can be made specific to a single species — the genes introduced into the tobacco were nematode-specific — whereas a cry protein affects a whole insect group. A sequence unique to the target parasite simply has no match in any other organism's transcriptome.

A fair qualification on both. Both require the pest to ingest the plant, so neither touches sap-suckers or organisms that do not feed on the crop. And the chapter's own caution applies to both: genetic modification of organisms can have unpredictable results when such organisms are introduced into the ecosystem, which is why the GEAC exists.