ABSTRACT

Glucosinolate accumulation in vegetative tissues of rape is a dynamic process, influenced by stresses such as tissue damage by herbivores and fungal infection. In addition to localised increases in glucosinolates, there is also a systemic “induction” following the local damage. The extent of such increased accumulation varies considerably between rape cultivars, and between Brassica species. There appears to be a correlation between the speed and extent of such “induction” and resistance to pathogens such as Sclerotinia sclerotiorum, such that the glucosinolate response could be a good marker for resistance. In contrast, certain glucosinolates appear to be necessary for clubroot (Plasmodiophora brassicae) infection of rape and other plants, and increased indolyl or aromatic glucosinolate content is associated with successful infection. Not all glucosinolates are equally important and effective - clubroot infection of Reseda alba may be restricted by high glucobarbarin/low indolyl content in the roots of this species. Other pest and pathogen interactions are similarly related to individual glucosinolates (or a narrow range). Flea beetle feeding markedly enhances the indolyl glucosinolate content of leaves, and these compounds also stimulate beetle feeding when added to artificial diets.

INTRODUCTION

The glucosinolate/myrosinase system is an effective defence against many pests and pathogens of Brassica crops, yet is exploited by certain specialist organisms to help them recognise and feed on or infect their host plants (see Bartlet 1996, Wallsgrove et al. 1998). Reducing or removing glucosinolates from rape tissues may reduce predation by specialists, but leaves the plants open to increased attack from many other organisms (e.g. Glen et al. 1990). So it is important to understand fully the relationship between glucosinolate content and host recognition and attack by specialist pests and diseases, so that appropriate glucosinolate profiles can be bred or engineered into the crop. It has become very clear recently that the glucosinolate/myrosinase system is dynamic, responding to environmental changes and to plant damage. Pre-treatment with elicitor compounds, which stimulate glucosinolate accumulation (Kiddle et al. 1994, Doughty et al. 1995a), can enhance resistance of the plant to subsequent infection by pathogens (Doughty et al. 1995b).

We have investigated the dynamics of glucosinolate accumulation, both locally and systemically, and attempted to relate this to disease resistance and insect herbivory. The role of glucosinolates in determining the host range of clubroot (Plasmodiophora brassicae) has also been studied. The importance of the wound/stress stimulated response has become clear, though factors other than the glucosinolates are also important.

MATERIALS & METHODS

The plant material used, and the growth conditions, were as described in Li et al. 1999a, and Ludwig-Mueller et al. 1999. Culture of the Sclerotinia isolate, and the infection and disease scoring methods were as described in Li et al. 1999b. Glucosinolates were extracted and analysed (as the desulphoglucosinolates) as described in Porter et al. 1991. Infection with Plasmodiophora, and analysis of fungal development, were as described in Ludwig-Mueller et al. 1999. Behavioural and feeding studies with flea beetles and slugs were as described in Bartlet et al. 1999.

RESULTS

Glucosinolates and Sclerotinia

The constitutive glucosinolate content of oilseed rape leaves is not correlated with either the seed content, or the susceptibility of the plant to Sclerotinia. Detailed survey of a range of Chinese and European rape cultivars revealed significant variation in leaf glucosinolate content and spectrum (Table 1). This variation was not related to field-assessed disease resistance.

Table 1. Variation in leaf glucosinolate content in rape cultivars

Double high = high seed glucosinolates + high seed erucic acid

Stem and root concentrations and profiles were similarly unrelated to seed content, or indeed to leaf content (Li et al. 1999a). The highest and lowest stem contents were found in two “double high” Chinese lines. There was a minor trend in 00 lines towards lower leaf content of alkenyl glucosinolates, but there were still exceptions to this.

Much more significant differences were found in the response of the various lines to infection with Sclerotinia. Amongst the European lines, the highly resistant cultivar Bienvenu showed marked local and systemic responses. In infected 3rd leaves there was a 10-20 fold increase in aromatic and indolyl glucosinolates, and a systemic increase in all classes of glucosinolate in the 7th leaf (by 26-99%). In the susceptible lines Cobra and Capricorn, the third leaves had a similar localised increase in glucosinolates, but systemic effects were much lower. Indeed, in Cobra the glucosinolate content of 7th leaves decreased by 20% following inoculation of the 3rd leaf. A similar pattern of strong systemic increases in resistant lines and weak or zero systemic increase in susceptible lines was found with the Chinese cultivars (Li et al. 1999b).

A second infection, on the 7th leaf of plants originally inoculated on the 3rd leaf, produced an additional response. In a resistant line, the second infection resulted in a marked increase in glucosinolate content in 7th and 10th leaves. Pre-inoculation of the 3rd leaf also reduced disease symptoms on the 7th leaf on reinoculation. A susceptible line responded only weakly to the second infection, and indeed glucosinolate content of the 10th leaf markedly decreased. Pre-inoculated plants of this line were not significantly protected against the second inoculation.

Glucosinolates and clubroot disease

Inoculation of a variety of glucosinolate-containing and non-glucosinolate-containing species with Plasmodiophora brassicae only produced disease symptoms or related responses in those species which contained glucosinolates, with the exception of Beta vulgaris, which also supported fungal development. In addition to the Brassica species studied, the pathogen developed (albeit more slowly) on roots of Tropaeolum majus and Carica papaya. On roots of the glucosinolate-containing Reseda alba no macroscopic symptoms developed, but young sporangia could be detected.

T. majus and C. Papaya both contain benzylglucosinolate, and no other glucosinolate. In addition, C.papaya also contains cyanogenic glucosides, one of very few species known to contain both classes of secondary metabolite (Bennett et al. 1997). There were local changes in glucosinolate content in both species, and systemic increases in T.majus. In C.papaya there is a localised decrease in cyanogenic glucosides after infection, but a marked systemic increase (Table 2).

Table 2. Changes in glucosinolate and cyanogenic glucoside content after clubroot infection

In roots of Plasmodiophora-infected R.alba there was a general decrease in glucosinolate content (-6 to -40%, depending on the compound). Over 80% of the glucosinolate content of the roots is made up of 2-OH-2-phenylethylglucosinolate, in the variety we were using, and the indolyl glucosinolates are generally low, in contrast to the roots of Brassica species where indolylglucosinolate are a major component (see Li et al. 1999a).

Rape responses to herbivore damage

Exposure of rape leaves to crucifer-specialising flea beetles (Psyliodes crysocephala) results in a systemic increase in the indolylglucosinolate content of the leaves, both young and old. Other glucosinolates were largely unaffected. The response seen is very similar to that previously decsribed for plants treated with methyl jasmonate (MJ) (Doughty et al. 1995a). Feeding tests of glucosinolate fractions from MJ-induced plants and control plants demonstrate that the induced glucosinolate profile is significantly more stimulatory to feeding. However, in whole plant tests these insects did not feed more on induced plants. In a similar set of experiments the generalist herbivore Deroceras reticulatum (grey field slug) fed significantly less on induced plants, and the altered glucosinolate content may have been partly responsible for this.

DISCUSSION

Glucosinolates play an important role in the interactions of rape with other organisms. They are significant for host plant detection by specialist herbivorous pests, and are involved in infection and development by fungal pathogens; but of course they are also major deterrents and protective agents against most non-specialist pests and pathogens. There appears to be no simple correlation between the constitutive glucosinolate content (or profile) and attraction/deterrence - rather it is the elicited response which is most significant. It is far from clear to what extent the glucosinolate contribution to induced defences is significant or important, but the speed and extent of the glucosinolate response would seem to be a good marker for effective responses by the plant. So far as Sclerotinia is concerned, almost all the resistant lines we have examined show a strong systemic increase in leaf glucosinolates; none of the susceptible lines have a comparable response. We had previously shown that prior elicitation of glucosinolates conferred a measure of resistance to two other rape pathogens, Peronospora parasitica and Alternaria brassicae (Doughty et al. 1995b). The abiotic elicitors used in that study, methyl jasmonate and salicylic acid, also induce other defensive responses, so again we cannot be sure of the precise role of glucosinolates in the overall interaction. Nonetheless, glucosinolate responses would seem to be an effective, and easily measured, indication of likely disease resistance.

Flea beetle feeding also produces an increase in glucosinolates, primarily the indolyl compounds. Although specialist flea beetles preferred the induced glucosinolate profile, when presented in an artificial medium, they did not feed more on induced plants. Thus other elements of the induced response must have a negative effect on flea beetle feeding. Other elements of the induced response must also reduce feeding by slugs, since they fed significantly less on induced plants and this was not solely attributable to altered glucosinolate content and profile. These other feeding- and MJ-induced plant defences are being investigated further.

Clubroot infection seems to be strongly influenced by glucosinolates in the root, yet it is clear that not all glucosinolates are equally important. Plasmodiophora prefers hosts with indolylglucosinolates as the major component of the root glucosinolate profile, though other aromatic glucosinolates may be important. Indolylglucosinolates are converted by the fungus to IAA. Successful colonisation of T.majus and C.papaya may be a consequence of conversion of benzylglucosinolate to phenylacetic acid (PAA), which has weaker auxin activity. In the case of R.alba, which is a poor host despite a high glucosinolate content in the roots, the role of 2-OH-2-phenylethylglucosinolate may be important, as it gives rise to a toxic, non-volatile phenyloxalidinethione (Kjaer and Gmelin 1957). This hypothesis needs to be tested, as we have subsequently found significant variation in the content of this particular glucosinolate in R.alba cultivars.

Much more work is required to understand the interactions between pest and disease organisms and the individual glucosinolates in a given species or cultivar. It is equally important to understand the stimulation of glucosinolate accumulation - what compounds are made in which tissues, where in the tissue are the “new” glucosinolates made and stored? The spatial distribution of glucosinolates in vegetative tissues is poorly understood, yet is likely to be highly significant in pest and disease responses. Work on this topic is currently underway at IACR, using some novel approaches. Other studies will be examining changes in enzyme activities in response to pests and pathogens, and induction of relevant genes (when the latter have been properly identified - Arabidopsis genome sequencing should eventually reveal all the glucosinolate biosynthesis genes).

Knowing the glucosinolate content of one part of the rape plant provides no useful information on the content or profile of other tissues on the same plant (see for example Li et al. 1999a). Nor is it a useful guide to the defensive responses of the plant. The glucosinolate/myrosinase system is much more dynamic than is often assumed, and this responsiveness plays a key role in the interaction of rape with other organisms.

REFERENCES

1. Bartlet E (1996). Agricultural Zoological Reviews 7, 89-116.

2. Bartlet E, Kiddle G, Martin A, Williams I and Wallsgrove RM (1999). Entomologia Experimentalis et Applicata (in press).

3. Bennett RN, Kiddle G and Wallsgrove RM (1997). Phytochemistry 45, 59-66.

4. Doughty KJ, Kiddle G, Pye BJ, Wallsgrove RM and Pickett JA (1995a). Phytochemistry 38, 347-350.

5. Doughty KJ, Bennett RN, Nashaat NI, Schrijvers S, Kiddle G, Pye BJ, Mitchell SE and Wallsgrove RM (1995b). In: Rapeseed Today and Tomorrow, GCIRC, pp 971-973.

6. Glen DM, Jones H and Fieldsend JK (1990). Annals of Applied Biology 117, 197-208.

7. Kjaer A and Gmelin R (1957). Acta Chemica Scandanavia 11, 906-907.

8. Li Y, Kiddle G, Bennett RN, Doughty K and Wallsgrove RM (1999a). Annals of Applied Biology 134, (in press).

9. Li Y, Kiddle G, Bennett RN and Wallsgrove RM (1999b). Annals of Applied Biology 134, (in press).

10. Ludwig-Mueller J, Bennett RN, Kiddle G, Ihmig S, Ruppel M and Hilgenberg W. (1999). New Phytologist 141, 443-458.

11. Porter AJR, Morton AM, Kiddle G, Doughty KJ and Wallsgrove RM (1991). Annals of Applied Biology 118, 461-467.

12. Wallsgrove RM, Doughty K and Bennett RN (1998). In: Plant Amino Acids - Biochemistry & Biotechnology (ed. Singh BK), Marcel Dekker, NY. pp. 523-561.