An arsenal of glucosinolates to defend winter-cress

Winter-cress (Barbarea vulgaris) is one of the many plant species to use repellent chemical compounds to defend itself against its attackers. It does so by producing glucosinolates – compounds that occur in many cruciferous plants, such as cabbage, mustard, and the species geneticists use as a model: Thale cress (Arabidopsis thaliana). Together with the enzyme myrosinase, glucosinolates form a two-component system: the enzyme and the glucosinolates are stored in the plant separately, but when the foliage is damaged they come together and the glucosinolates are converted by the enzyme.

As the glucosinolates and breakdown products are often toxic to generalist herbivores, they reduce herbivore damage to plants. Specialist herbivores, however, use these compounds to identify suitable host plants. The glucosinolates attract these specialist herbivores to the plant and stimulate them to eat it. The compounds formed by the enzyme-induced chemical conversion of the glucosinolates are also responsible for the taste of the cabbage and mustard species that humans like to eat. There are over 100 types of glucosinolate, each with its own distinctive chemical structure which determines the breakdown product that is formed when the plant is damaged. And each plant species has its own characteristic assemblage of glucosinolates.

In her PhD research, Hanneke van Leur investigated whether the differences in glucosinolate profile also influence plant resistance to herbivores. When she examined populations of winter-cress (Barbarea vulgaris) in the Netherlands, she found that not all the plants synthesised exactly the same glucosinolates. Most of them synthesise glucobarbarin, a glucosinolate typical of this species – hence its name. Winter-cress populations in Germany, Belgium, France and Switzerland were also found to consist solely of plants in which glucobarbarin was mainly found. In half of the Dutch populations, a small proportion of the plants (2-22%) synthesised another glucosinolate, called gluconasturtiin. The difference in the chemical structure of these two glucosinolates is very small: glucobarbarin has just one more hydroxyl group than gluconasturtiin. Nevertheless, in biological terms this can be critical: if gluconasturtiin comes into contact with myrosinase, a toxic, strong-tasting isothiocyanate is formed, whereas in plants containing glucobarbarin, oxazolidinethiones are created, because of the position of the hydroxyl group. It is not known whether these oxazoldinethiones are very toxic, though we do know that if ingested by mammals they inhibit the uptake of iodine, which may lead to thyroid problems. So, the winter-cress glucosinolate polymorphism consists of two chemotypes. Van Leur describes these chemotypes in her PhD thesis, and exploits the polymorphism in order to study the effects of different glucosinolates on herbivores. 

Winter-cress	Caterpillar of the cabbage moth on Winter-cress

The difference in the glucosinolate profile is consistently present in all the organs of winter-cress, but is more pronounced in the aboveground parts of the plant than in its roots. The glucosinolate profile does not change when the plants are induced naturally by insects, or artificially by the addition of jasmonic acid. By crossing plants, Van Leur demonstrated that the ability to synthesise glucobarbarin is hereditary and is regulated by a dominant gene. Several candidate genes were identified, on the assumption that a specific enzyme is responsible for the hydroxylation of gluconasturtiin to glucobarbarin. Further research is needed to establish whether one of these candidate genes is actually responsible for the difference between the chemotypes.

When Van Leur investigated what effect the difference in chemotype has on leaf and root herbivores, she found that the plants containing glucobarbarin were very resistant to the generalist leaf-eating caterpillars of the cabbage moth (Mamestra brassicae). Only a few caterpillars survived on plants containing glucobarbarin; when given the choice, they showed strong preference for eating plants containing gluconasturtiin. The female moths nonetheless laid about the same number of eggs on leaves of both chemotypes. The caterpillars of the specialist small cabbage white butterfly grew equally well on both types of plant and in the choice test did not prefer one above the other. However, the larvae of the specialist cabbage root fly did worse on the roots of the gluconasturtiin type than on the roots of the glucobarbarin type. Measurement of the impact on the plant of an infestation of root fly revealed that root fly infestation halved the root and shoot mass of the plants of both chemotypes and decreased the concentrations of nutrients such as sugars and amino acids.

A field trial was conducted, so that the greenhouse results could be more reliably extrapolated to the field. It entailed planting plants of both chemotypes in trial plots and then counting the number of herbivores on them every week, for two years. Some aboveground insect species showed a preference for a certain chemotype; others did not. Small cabbage white butterflies preferentially laid their eggs on plants containing gluconasturtiin, but specialist flea beetles and gall midges were more numerous on plants containing glucobarbarin. The cabbage aphid and green peach aphid showed no preference: they were equally abundant on both chemotypes. Three to four times a year, Van Leur dug up a sample of plants so she could analyse the root herbivores and soil nematodes. She found no difference between the chemotypes in the composition of these soil-dwelling organisms.

Finally, Van Leur performed extensive metabolomic analyses using liquid chromatography time-of-flight mass spectrometry, in order to study the chemical differences between the two chemotypes more closely. Multivariate analyses revealed that glucosinolates are indeed responsible for the biggest chemical difference between the two types; the difference was more pronounced in the shoots than in the roots. In addition to the glucosinolates, Van Leur found eight previously unidentified compounds that were present in small amounts and differed between the chemotypes. Well-known defence compounds such as flavinoids and sapinoids were identified, but their concentrations did not differ between the chemotypes. This makes it very likely that the differences in growth and herbivore preference are primarily the result of the differences in glucosinolate profile.

Conclusions
From the results of the experiments in the greenhouse and in the field it can be concluded that the structure of the glucosinolates are responsible for the significant differences in resistance against different herbivores. However, one chemotype is not more resistant than the other in all cases. Which of the two chemotypes is more advantageous therefore depends on which herbivores occur in a given plant population. This accounts for why the glucosinolate polymorphism is maintained in natural populations.