A second major area in which plant-derived natural products have been linked with defence is concerned with invasive plants and the rather controversial area of allelopathy. Invasive plants include species such as wild oat (Avena fatua) (Fig. 5a), a persistent weed of cereal crops; and spotted knapweed (Centaurea maculosa), an exotic species that is aggressively displacing native plants in the western USA (Fig. 5b). Plants like these pose a threat to agricultural systems and natural environments throughout the world. Establishing the factors that determine invasiveness is not trivial, as resource availability, the physical environment and interactions with native plants, animals and microbes (both detrimental and beneficial) will all have an impact on competitiveness (Shea & Chesson, 2002; Fitter, 2003). Invasive species may be particularly successful at establishing and spreading in alien environments because they have escaped competition from their natural neighbours. However, it has also been argued that the ability to produce allelopathic chemicals may contribute to their success (Whittaker & Feeney, 1971; Rice, 1984; Williamson, 1990; Callaway & Aschehoug, 2000; Inderjit & Duke, 2003). At this point it may be helpful to consider some definitions, as interpretations of the term ‘allelopathy’ vary.
Figure 5. Invasive plant species. (a) Wild oats (Avena fatua) invading a wheat crop (photograph courtesy of the Agricultural Development and Advisory Service, Boxworth, Cambridge, UK). (b) Spotted knapweed (Centaurea maculosa), an exotic species that is aggressively displacing native plants in the western USA (photograph courtesy of Ray Callaway, University of Montana, Missoula, MT, USA).
Download figure to PowerPoint
1. What is allelopathy?
‘The described phenomenon that one plant can influence another plays an important role in physiology, so it deserves an appropriate term. For this I coin the word allelopathy from the Greek words “allelon” meaning mutual and “pathos” meaning harm or affection.’
Note that this definition encompasses both positive and negative effects. The definition of allelopathy was subsequently broadened further by Rice (1984) to cover the effects of plants and their associated microorganisms on growth of other plants through the release of chemical compounds into the environment. As Inderjit & Duke (2003) point out, this extends the term to include almost all aspects of chemical ecology. The current working definition of the term ‘allelopathy’ tends to focus specifically on the negative effects of plant compounds – the suppression of the growth of neighbouring plants by the release of toxic compounds into the environment (Baldwin, 2003; Fitter, 2003; Inderjit & Duke, 2003).
2. Release of phytotoxic compounds from plants
Phytotoxic (potentially allelopathic) chemicals can originate from aboveground parts of plants and be released into the environment through foliar leaching, volatilization, residue decomposition and debris incorporation. However, many phytotoxic compounds of plant origin are produced by roots and are released directly into the soil (Inderjit & Duke, 2003). As emphasized previously, roots of different plant species produce a diverse array of natural products, and these compounds are often organ- and/or species-specific (Flores et al., 1999; Inderjit & Duke, 2003; Narasimhan et al., 2003; Walker et al., 2003a; Tan et al., 2004; D’Auria & Gershenzon, 2005; Bais et al., 2006). The production and/or release of these root-derived natural products is commonly constitutive, but may be induced by biotic or abiotic stresses (Lydon & Duke 1993; Bais et al., 2002; Bertin et al., 2003). Mechanisms of release of biologically active compounds from roots are not well understood, but are likely to include both diffusion and controlled release via transporters and vesicles (Neumann & Römheld, 2001; Bertin et al., 2003; Walker et al., 2003a). Improved methods for analysis of the natural product content of root exudates in Arabidopsis and other plant species are now opening up new opportunities for identification of root-derived compounds and for investigation of their synthesis, regulation, release and function (Fan et al., 1997; Narasimhan et al., 2003; Walker et al., 2003a, 2003b; Steeghs et al., 2004; Tan et al., 2004; Bais et al., 2006).
Various crop plants and weeds are known to release phytotoxic compounds into the soil at bioactive concentrations (Niemeyer, 1988; Minorsky, 2002; Inderjit & Duke, 2003). Once outside the plant, the biological activity of chemicals that are released into the rhizosphere may be altered by chemical oxidation, microbial degradation or immobilization by irreversible binding to soil particles (Bertin et al., 2003). Assessment of the effective concentration of individual phytotoxic chemicals in the rhizosphere, and of the impact of these compounds on plants, therefore represents a substantial challenge (Weidenhamer, 2005). To date there are few cases in which specific compounds have been identified as allelopathic agents (Duke et al., 2001), and there are no definitive tests of the contribution of specific compounds to the ability of invasive plant species to suppress the growth of susceptible plants (Baldwin, 2003). Nevertheless, considerable progress has been made in various aspects of allelopathy research. Some examples of this are outlined below.
3. Allelopathy in cereals
Within the past 50 yr, weed management has become almost completely dependent on herbicides (Duke et al., 2001). The development of crops with enhanced allelopathic properties, through traditional breeding or genetic modification, has potential for controlling weeds and improving yield and may represent a promising alternative to the application of chemicals (Duke et al., 2001; Minorsky, 2002). Examples of natural products that have been implicated in allelopathy in cereals include the benzoxazinoids, defence-related compounds that occur constitutively as glucosides in certain members of the Gramineae and in some dicots. The primary hydroxamic acid in rye is 2,4-dihydroxy-1,4-benzoxazin-3-one (DIBOA), while its methoxy derivative 2,4-dihydroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA) (Fig. 2c) is predominant in maize and wheat (Barnes & Putnam, 1987; Niemeyer, 1988; Sicker et al., 2000; Wu et al., 2001a). These compounds have also been associated with chemical defence against pests and pathogens (Niemeyer, 1988; Gierl & Frey, 2001). The allelochemical quinone sorgoleone (Fig. 2d) is released from sorghum roots (Duke et al., 2001), while the diterpene momilactone B has been proposed to act as an allelopathic compound in rice (Fig. 2e) (Kato-Noguchi et al., 2002). This list is by no means exhaustive, and a variety of other compounds including phenolics have also been implicated in allelopathy in these and other cereals (Minorsky, 2002; Inderjit & Duke, 2003; Kong et al., 2004). Oats are particularly invasive when compared with other cereals. Phenolics and l-tryptophan have been identified as potential allelopathic agents in this cereal (Guenzi & McCalla, 1966; Guenzi et al., 1967; Schumacher et al., 1983; Pérez & Ormeño-Nuñez, 1991; Kato-Noguchi et al., 1994a, 1994b; Fay & Duke, 1977). The antimicrobial triterpene glycoside avenacin A-1 (Fig. 2b) is also released from oat roots into the soil (Carter et al., 1999). Triterpene glycosides have a diverse range of biological activities, and may contribute to allelopathy (Hostettmann & Marston, 1995).
Characterization of the genes and enzymes required for the synthesis of compounds that are implicated in allelopathy should allow the biosynthetic pathways to be manipulated, so enabling direct tests of the contribution of these compounds to suppress the growth of weeds and other plants in laboratory-, glasshouse- and field-based experiments. It may then be possible to use this knowledge to enhance production of allelochemicals in plants that already produce them, or to transfer the ability to synthesize these allelopathic compounds into other important crop species (Duke et al., 2001; Bertin et al., 2003; Dayan et al., 2003; Wilderman et al., 2004; Xu, M. et al., 2004; Yang et al., 2004a, 2004b). The genes for the pathway leading to the synthesis of DIBOA and DIMBOA have been cloned from maize (Frey et al., 1997; Sicker et al., 2000; Gierl & Frey, 2001). Corresponding mutants are available for a number of these genes, although the effects of mutation on the ability to suppress other plants have not been reported. It is intriguing that the genes for the DIBOA/DIMBOA and avenacin biosynthetic pathways exist as module-like clusters in maize and oats, respectively (Frey et al., 1997; Papadopoulou et al., 1999; Haralampidis et al., 2001; Osbourn et al., 2003; Qi et al., 2004). Terpene synthases involved in the synthesis of rice momilactones also exist as functional clusters (Wilderman et al., 2004). The reasons for this clustering are unclear. Mechanisms that act to disperse genes (translocation, inversion and unequal crossing over) are well known in eukaryotes, and genes associated with other well characterized secondary metabolic pathways, such as anthocyanin biosynthesis, are generally unlinked. Horizontal transfer of gene clusters from other organisms cannot be the sole explanation for the existence of the clusters, as there is clear evidence of recruitment of cluster components from plant primary metabolism (Gierl & Frey, 2001; Qi et al., 2004). It seems that, in some cases, physical clustering of genes for the synthesis of protective chemicals is favoured because it will enable pathway genes to be inherited as a functional unit (Qi et al., 2004). Clustering also has the potential to facilitate coordinate regulation of gene expression at the chromatin level, a factor that may be important for stringent regulation of the synthesis of potentially phytotoxic plant compounds (Qi et al., 2004). The intriguing question of how these gene clusters came into being remains to be addressed. Regardless of this, the idea of transferring these chemical defence clusters between species via classical or GM-based approaches is tantalizing.
Germplasm of various crop plants, including barley, wheat, rice, sorghum, maize and oats, has been screened for allelopathic potential, and significant variation has been identified in each of these crops (Lovett & Hoult, 1995; Nimbal et al., 1996; Fay & Duke, 1997; Wu et al., 2000, 2001b; Minorsky, 2002; Olofsdotter et al., 2002; Sánchez-Moreiras et al., 2003). Although commercial crops with enhanced allelopathic potential have not yet been developed, these findings provide opportunities for improving allelopathy in rice through conventional breeding strategies by using marker-assisted selection.
A general caveat associated with the development of crop plants with enhanced allelopathic traits is that they are likely to be more competitive in natural ecosystems, although it may be possible to circumvent this problem by employing appropriate agronomic practices in order to minimize the risk of escape (Duke et al., 2001).
4. Sorghum and sorgoleone
Sorghum leaves synthesize and secrete red pigmented anthocyanidins as part of the induced defence response (Section II). The roots of sorghum produce a different pigmented natural product, the quinone sorgoleone (2-hydroxy-5-methoxy-3-[(8′z,11′z)-8′,11′,14′-pentadecatriene]-p-benzoquinone) (Fig. 2d), which is exuded from the root hairs in golden brown droplets (Nimbal et al., 1996; Czarnota et al., 2001; Czarnota et al., 2003) (Fig. 6a). Synthesis of sorgoleone is tissue-specific and occurs only in the root hairs (Czarnota et al., 2001; Yang et al., 2004b). Sorghum species are often used as cover crops because of their ability to suppress weeds, and evidence suggests that sorgoleone may contribute to this suppressive activity (Duke et al., 2001). Sorgoleone has potent phytotoxic activity and inhibits the germination and growth of susceptible weeds at concentrations as low as 10 µm (Einhellig & Souza, 1992; Nimbal et al., 1996). Root hairs can represent up to 80% of the root surface of cultivated crops, so release of sorgoleone into the rhizosphere is likely to be favoured (Dittmer, 1937; Yang et al., 2004b). Indeed, sorgoleone concentrations in soil under sorghum crops can easily reach 10–100 µm (Netzley et al., 1988). The compound has been reported to have a variety of effects on plant metabolism, including inhibition of photosynthesis and respiration (reviewed by Inderjit & Duke, 2003; Hejl & Koster, 2004).
Figure 6. Secretion of sorgoleone. (a) Droplets containing sorgoleone exuding from root hairs. Bar, 50 µm (b,c) Transmission electron micrographs showing longitudinal sections of Sorghum root hairs. Root exudates (stained black) are deposited between the plasma membrane and the cell wall. Bars, 1 µm. CW, cell wall; ER, endoplasmic reticulum; Mt, mitochondria; S, root exudates; So, root exudates at various stages of secretion; V, vacuole. Reproduced from Czarnota et al. (2003), with permission of University of Chicago Press.
Download figure to PowerPoint
Characterization of the sorgoleone biosynthetic pathway and cloning of the cognate genes are currently under way (Duke et al., 2001; Dayan et al., 2003; Yang et al., 2004a, 2004b). The precise mechanism of secretion of sorgoleone is not understood, although a detailed study of the anatomy of sorgoleone-secreting root hairs suggests that this process is likely to involve vesicle trafficking (Czarnota et al., 2003). Transmission electron microscope studies suggest that the compound is synthesized in association with the ER, transported through the root-hair cells in lipophilic vesicles, then deposited between the plasma membrane and the cell wall (Czarnota et al., 2003) (Fig. 6b,c). Although these studies did not involve specific detection of sorgoleone, this compound and its analogues comprise approx. 90% of the content of root hair exudate. It is therefore reasonable to assume that the globules that are visible in Fig. 6b,c contain sorgoleone.
5. Spotted knapweed and (–)-catechin
Recent data from work with spotted knapweed (C. maculosa) has provided a solid base of evidence linking production of a specific root-derived compound with allelopathy, and has made inroads into addressing the mechanism of toxicity of this compound to susceptible plants. Vivanco and coworkers have identified the flavonol (–)-catechin (Fig. 2f) as the primary phytotoxic compound that is released into the soil from the roots of spotted knapweed (Bais et al., 2002, 2003). Although spotted knapweed roots exude a racemic mix of (±)-catechin, only (–)-catechin is phytotoxic. Interestingly, (+)-catechin (but not (–)-catechin) has antimicrobial activity towards root-infecting pathogens, leading to speculation that the enantiomers may have complementary functions in root protection (Bais et al., 2002); the observation that (±)-catechin also appears to play an autoinhibitory role in regulating conspecific seedling establishment is intriguing (Perry et al., 2005). The mechanism of release of these compounds into the soil is not yet known.
Recent experiments have provided insights into the mechanism of phytotoxicity of (–)-catechin. (–)-Catechin inhibits the germination and growth of other sensitive plant species at concentrations that occur naturally in soils, and causes cell death when applied to roots of Arabidopsis and the sensitive knapweed species Centurea diffusa at concentrations of 100 µg ml−1 (300 µm). Further analysis has revealed that (–)-catechin-induced cell death is preceded by production of ROS and a transient elevation in Ca2+ levels (Bais et al., 2003). These data suggest that ROS-induced Ca2+-dependent signal transduction triggers cell death. The tools and resources available for Arabidopsis now offer an opportunity to dissect the mechanism of (–)-catechin-induced cell death in more detail.
6. Living together
Plant species that normally coexist together are likely to have developed mechanisms of tolerating each other's toxins (Fitter, 2003). Examples are known in which resistant species detoxify allelochemicals by oxidation, carbohydrate conjugation or sequestration (Wieland et al., 1998; Sicker et al., 2001; Inderjit & Duke, 2003). Weir et al. (2006) suggest that resistance to the spotted knapweed allelochemical (±)-catechin may be conferred by increased secretion of oxalate, which is believed to protect against damage incurred by ROS. A number of allelochemicals have toxicity rivalling that of synthetic herbicides (Weidenhamer, 2005). As resistance in plants evolves rapidly in response to man-made herbicides (Powles & Holtum, 1994), resistance to allelochemicals of exotic invaders may therefore also be expected to evolve rapidly in native plant populations (Callaway et al., 2005). Unravelling the significance and consequences of production of allelochemicals by plants, and putting this into the broader context of plant competition and plant–microbe interactions, represents a substantial challenge for the future.