- Top of page
- Materials and Methods
Almost all plant species that have been studied respond to phytophage attack by changing their levels of chemical compounds. In many cases, these induced responses have been found to increase plant resistance to subsequent attacks by the same or other plant enemies (Karban & Baldwin, 1997). So far, studies on induced responses have mainly focused on the aerial parts of the plant and on how these responses affect above-ground herbivores and pathogens (van Dam et al., 2003). This focus on the shoot ignores not only the impact that below-ground phytophages may have on plant survival and plant biodiversity (van der Putten et al., 2001; de Deyn et al., 2003), but also that root damage may induce defensive responses as well (Birch et al., 1990; Ludwig-Müller et al., 1997). Many induced responses are systemic throughout the plant (Baldwin et al., 1994; van Dam et al., 2001; van Dam et al., 2003), so that shoot feeders may alter food quality for root feeders, and vice versa (Bezemer et al., 2003; Wäckers & Bezemer, 2003). As plants are attacked by both root and shoot phytophages, interactions between above-ground and below-ground induced defences are likely to occur.
There are several ways in which above-ground and below-ground induced defences can interact. First, simultaneous induction in roots and shoots may result in competition for limited resources, biosynthetic capacity, or defence compounds, especially when damage alters the relative sink-strength of plant organs (van Dam & Vrieling, 1994). Alternatively, above-ground and below-ground induced responses may interact when they are elicited by phytophages that trigger different signalling pathways in the plant. It has been well documented in above-ground studies that the salicylic acid (SA) signalling pathway, involved in induced responses against pathogens (Hammerschmidt & Smith-Becker, 1999), may suppress jasmonic acid (JA)-induced responses elicited by herbivore feeding (Creelman & Mullet, 1997; Felton et al., 1999; Preston et al., 1999; Thaler et al., 1999). In several plant species, both JA- and SA-induced responses have been found to be systemic throughout the plant (van Dam et al., 2001; Rostás et al., 2003). Therefore, JA–SA interactions may also occur between root and shoot induced responses.
In this study we use two naturally occurring Brassica species to evaluate interactions between shoot and root induced responses that are elicited by JA and SA application. Although Brassicaceae contain several classes of chemical compounds that may serve as defences, such as protease inhibitors (Cipollini & Bergelson, 2000; De Leo et al., 2001), saponins (Shinoda et al., 2002) and anthocyanins (Rostás et al., 2002), the most well-known defence compounds in this plant family are the glucosinolates (GS; Halkier & Du, 1997, Fahey et al., 2001). Based on the chemical structure of their side chain, the GS can be subdivided in different classes, such as aliphatic, aromatic and indole GS (Fahey et al., 2001). GS themselves may deter generalist herbivores (Li et al., 2000), but the hydrolysis products that are formed when cell rupture brings them into contact with myrosinase, an enzyme stored in specialised plant cells (Rask et al., 2000), are generally much more potent. Depending on the chemical structure of the GS side chain and the reaction conditions (e.g. pH), combination of the enzyme with the GS results in several different noxious and toxic products, such as isothiocyanates (ITC), oxazolidine-2-thiones and nitriles (Wittstock & Gershenzon, 2002).
We applied JA and SA solutions to roots and shoots of Brassica oleracea and B. nigra plants, which are two naturally occurring species from Western Europe. Although both species belong to the same genus, they are quite different in morphology, life-history and chemistry. B. oleracea is a spring perennial that has smooth, waxy leaves. The shoots of B. oleracea contain several different glucosinolates, among which the aliphatic gluconapin (3-butenyl GS) and the indole GS glucobrassicin are the most prominent (Mithen et al., 1987). B. nigra, on the other hand, is a summer annual, which has a rough appearance because its leaves are covered with trichomes (Traw & Dawson, 2002a). Next to the main GS sinigrin (2-propenyl GS) the plant species produces minor quantities of indole GS such as glucobrassicin (Traw & Dawson, 2002b,a). Moreover, when both plant species are grown under similar conditions in the glasshouse, the total GS levels in shoots of B. nigra are 3.5 times higher than those in shoots of B. oleracea (Harvey et al., 2003).
Based on previous studies in cultivated Brassica species, we expect that both JA and SA application will elicit a GS response and that these responses are systemic (Ludwig-Müller et al., 1997; Bartlet et al., 1999). Individual classes of GS, however, may respond differently to induction treatments (Mikkelsen et al., 2003). Indole GS, for example, consistently increased after application of JA or methylJA to shoots of the oilseed rapes Brassica napus and B. rapa, the mustard B. juncea or the Chinese cabbage B. campestris cv. pekinensis (Bodnaryk, 1994; Doughty et al., 1995; Ludwig-Müller et al., 1997; Bartlet et al., 1999). The levels of aromatic GS increased as well only in B. campestris, whereas in the other species the levels of aliphatic and aromatic GS levels remained unchanged after JA treatment (Doughty et al., 1995; Ludwig-Müller et al., 1997; Bartlet et al., 1999). By contrast, SA application to the roots increased all classes of GS in the shoots of B. napus, but aromatic GS levels increased more than indole or aliphatic GS levels (Kiddle et al., 1994). Because the structure of GS is closely related to their efficacy as defences against different phytophages (Brown & Morra, 1997; Potter et al., 1999), differential induction of GS indicates that plants are able to specifically tailor their response to the enemy that is feeding. Interactions between simultaneous induction events in roots and shoots may interfere with the optimisation of the response, and eventually affect the amount of damage to the plant.
By applying JA and SA solutions to roots and shoots in all possible combinations, we specifically addressed the question whether application of JA or SA to one organ elicits a systemic response in the other organ, and whether these responses interact. By analyzing the different classes of GS, we were able to study whether JA and SA applications trigger differential induction of specific classes of GS. In addition to levels of total and individual classes of GS, we also analyzed changes in biomass, which may indicate potential costs involved in the production of inducible compounds.
- Top of page
- Materials and Methods
This study shows that root induction events, as mimicked by the application of JA, have a significant systemic impact on total shoot GS levels in two wild Brassica species. This indicates that root phytophages, which trigger the JA-signalling pathway, may significantly alter food quality for their above-ground feeding counterparts by changing the levels of defence compounds in the shoot. This may not only affect the performance of shoot phytophages on induced plants, but also determine how much damage will be done to the shoot (Bezemer et al., 2003). JA application to the shoot mainly increased local GS levels. In B. nigra, however, the average level of aromatic GS (in this case gluconasturtiin) in roots of JA-treated plants significantly increased to more than 8 µmoles g−1 dry mass. Potter et al. (1999) found that this is a critical threshold level for gluconasturtiin in B. napus roots to significantly reduce susceptibility to the generalist nematode Pratylenchus penetrans. The increase in aromatic GS was the strongest in plants that were treated with JA to their roots, but was also clearly present in shoot-treated plants. Thus even if total GS levels in roots are not significantly affected, shoot feeders that induce the JA signalling pathway in B. nigra may alter host-plant suitability for specific groups of root feeders, such as nematodes. As in B. oleracea, B. nigra plants treated with JA to both roots and shoots had higher total GS levels in their shoots than plants treated with a single application of JA only. Although the effect was not completely additive, it indicates that in both plant species the total GS response could have been even stronger than observed in this experiment.
We found that shoot and root treatments with similar quantities of JA result in similar increases of total GS, but significantly different GS profiles in B. oleracea shoots. Such a shift in GS profiles may be of great ecological importance, as the structure of a GS is closely related to its potency as defence to phytophages. Moreover, because application of jasmonates also may also increase the levels of the hydrolytic enzyme myrosinase in the leaves (Taipalensuu et al., 1997), the JA-induced differences in GS levels probably translate directly into different profiles of the biologically active hydrolysis products upon damage. The aliphatic GS that were induced after JA application to the roots, yield isothiocyanates, which are generally recognized as potent antimicrobial and antifungal compounds as well as deterrents of a wide range of generalist and specialist herbivores (Brown & Morra, 1997; Fahey et al., 2001; Lambrix et al., 2001; Agrawal & Kurashige, 2003). By contrast, the indole GS, which were induced after shoot application, produce isothiocyanates that are unstable and decompose spontaneously (Fahey et al., 2001). Because of this instability, indole GS are far less effective deterrents or toxins, and their presence may even attract certain herbivores to the plant (Chew, 1988; Moyes et al., 2000). These effects may also extend to third and higher trophic levels, because the isothiocynates of aliphatic GS are partly volatile and thus may be used as host-finding cues by predators and parasitoids that are specialised on hosts that feed on GS-containing plants (Brown & Morra, 1997; Bradburne & Mithen, 2000). Plants that were treated with JA to both organs simultaneously showed an induced GS profile that was similar to plants treated with JA to their shoots. This indicates that the shoot-induced response has priority over the root-specific response. The specific GS response of B. oleracea to root and shoot JA application was not altered by SA application to the other organ, as has been observed in above-ground induced responses in several other plant species (Karban & Baldwin, 1997). This does not completely preclude interaction between JA and SA signalling pathways: it was shown that exogenous application of SA reduces the expression of myrosinase transcripts (Taipalensuu et al., 1997). Myrosinase activity, however, was not measured in this study and moreover it is not known which level of myrosinase activity suffices to hydrolyze all GS at a wound site. Therefore, the outcome of these potential interactions can only be analysed at the phenotypic level, that is by measuring the volatile profile of damaged plants that were pretreated with JA and SA. In B. nigra plants, we did not find a significant difference in shoot GLS levels or patterns between the different JA treatments.
Unlike in cultivated Brassica species (Kiddle et al., 1994; Ludwig-Müller et al., 1997), SA applications to B. nigra and B. oleracea did not increase overall GS levels in shoots or roots. In both plant species, total root GS levels even decreased in plants treated with SA to the roots, especially in B. oleracea. This is consistent with the GS levels found in SA-over-expressing A. thaliana mutants, which also had lower GS levels than wild-type plants (Mikkelsen et al., 2003). A decrease in root GS synthesis would not necessarily decrease shoot levels as well, because above-ground and below-ground GS levels and profiles are regulated independently (Sang et al., 1984; Potter et al., 1999).
Additionally, we observed a combined morphological-chemical response to SA leaf application in B. nigra plants. The dark purple ring around the trichomes that were formed after treatment most likely was a result of accumulation of anthocyanins in cells surrounding the trichome bases. It is known from studies on other, cultivated, Brassica species, that herbivory as well as fungal infection of the leaves may increase shoot anthocyanin levels (Rostás et al., 2002). Trichome responses in B. nigra also occur after leaf damage by lepidopteran larvae, which increases trichome densities on younger leaves (Traw & Dawson, 2002a). This indicates that trichome responses are an intrinsic part of the induced response in this plant species.
Root application of SA to B. nigra, on the other hand, induced the formation of lesions on the leaves, which is known as the hypersensitive response, a common plant response to pathogen attack (Hammerschmidt & Nicholson, 1999). Because we did not observe any direct phytotoxic effects when SA was applied directly to the shoot, this response most likely is part of the systemic acquired resistance response, which is controlled by the SA signaling pathway (Pieterse et al., 2002). As the lesion formation is a systemic response, it shows that the lack of response in shoot GS levels after SA application is not caused by a lack of SA uptake or transport from roots to shoots.
Interestingly, in neither of the plant species was SA application found to depress the GS response to JA application, as has been reported in other plants species (Creelman & Mullet, 1997; Felton et al., 1999; Preston et al., 1999; Thaler et al., 1999). Similarly, simultaneous application of JA to roots and shoots did not decrease the total GS response in the shoot, probably because root and shoot biosynthesis are regulated independently. This indicates that the GS response in these plant species is not constrained when different phytophages attack the plant at the roots and at the shoots simultaneously. Brassicaceae, however, are known to produce many other inducible compounds that have defensive properties, and that are induced in response to different types of phytophages. Possibly, there are constraints on the simultaneous induction of different classes of compounds. The eventual effects of such interactions on naturally occurring root and shoot phytophages, as well as on the plant's reproductive output, remain to be assessed.
In our experiment we found no indications that increased GS levels following JA applications reduced plant biomass or growth within 1 wk after induction. This indicates that the enhanced production of these compounds per se is not costly to the plant in this short time period. However, we can not rule out that there may be other, ecological, costs to induction of GS and possibly other inducible compounds in these plants. These costs may only emerge under circumstances that are ecologically more realistic, for example when plants that are induced are in competition for nutrients and light (van Dam & Baldwin, 2001). Especially in B. nigra, which is a rapidly growing annual species found in dense stands of up to 150 individuals per square meter (J. Harvey, pers. obs.), such cost may be of crucial importance.
Our results have shown that there is a great potential for induced responses to mediate interactions between above-ground and below-ground phytophages feeding on wild Brassica species. This potential only increases if one realizes that next to GS and trichomes, there are several other defensive compounds found in these species that are inducible by herbivores and pathogens, for example protease inhibitors or volatile compounds that attract natural enemies (Geervliet et al., 1997; De Leo et al., 2001). We expect that in natural environments, where plants are constantly under attack by a wide range of root and shoot phytophages, such interactions may be of great ecological importance to plants, phytophages and their natural enemies alike.