For humans, a slightly bitter taste is the worst of the glucosinolates - and, in fact, these compounds are the ones thought to have cancer fighting activity - but for small insects who want to make their whole meal on raw Arabidopsis, the problem can be much more significant. The plant makes both a family of compounds called glucosinolates and an enzyme called myrosinase. A healthy leaf stores the two components in separate compartments. However, when a caterpillar or other herbivorous insect takes a bite out of the leaf, cell walls and compartments are broken down, allowing the enzyme and glucosinolates to come into contact, at which time the enzyme converts the bitter-tasting compound into a toxic one.

Of course, over evolution some pests have found ways to get around the problem. When Kliebenstein tested a pest who feeds on a wide range of plants and is, therefore, categorized as a generalist, he found that the caterpillars were sensitive to the gluocsinolates. However, when he tested a specialist pest that has evolved to feed almost exclusively on brassicas, he found that, across the board, they were resistant to the toxic effects of the glucosinolates.

So just how does a specialist pest, like Plutella xylostella, avoid the toxin? To get an idea of how the caterpillar was metabolizing the glucosinolates, Kliebenstein and his colleagues collected the insect's feces, dissolved it in water and analyzed it with high pressure liquid chromatography (HPLC). Interestingly, the team found that the glucosinolates were passing through the animal relatively unchanged, as if the plant didn't even have the myrosinase enzyme at all.

The researchers subsequently found that one of the major proteins in the caterpillar's spit is an enzyme that removes a sulfur group from the glucosinolates. Without this sulfur group, the plant's own enzyme, myrosinase, doesn't recognize the glucosinolate and therefore can't convert it to a toxic compound, leaving the pest free to chew away at the now defenseless plant.

"This is a nice bit of work," said Neal Gutterson, who works on functional genomics in Arabidopsis at Mendel Biotechnology, Inc. in Hayward, California, "especially the part about how the specialist detoxifies the compound."

Joseph Anderson, from the USDA/ARS Crop Production and Pest Control Research Unit at Purdue University in West Lafayette, Indiana, told BioMedNet News that Kliebenstein's work on glucosinolates in Arabidopsis is "quite applicable to other plant systems." Because many plants use secondary metabolites, like glucosinolates, what scientists learn in this model system will be generally transferable to other plants - and their pests.