Two-component plant defenses are better than one

Plants that store toxic metabolites as defenses against herbivores and pathogens must ensure that they do not poison themselves. One strategy is to store the toxin as an inactive glycoside and hydrolyze the sugar moiety on herbivore damage to release the active toxin. This scheme, referred to as a two-component defense, requires a glycoside protoxin, an activating glycosidase, and a spatial separation between the two components in the plant to prevent premature activation. Several well-known groups of plant toxins, including cyanogenic glycosides (1), glucosinolates (2) and benzoxazinoids (3), follow this pattern, with saponins having been recently added to this list (4). However, given that saponins typically have multiple sugar residues, it is often unclear which compounds are biologically active, how activation is accomplished, and how plants avoid autotoxicity. A new study by Han et al. on the triterpenoid saponins of the common ivy (Hedera helix) illuminates the two-component defense of this species in exquisite detail (5).

Common ivy produces two triterpene saponins with an identical hydrophobic aglycone core, except that α-hederin has a two-sugar side chain attached to C-3, while hederacoside C has a two-sugar side chain at C-3 plus a three-sugar side chain at C-28. These compounds have been studied for their chemical and pharmacological properties (6), but their role in a two-component defense system had not been appreciated. Han et al. demonstrated that tissue damage to common ivy leads to a decrease in hederacoside C levels and a concomitant increase in α-hederin levels over a short time span, consistent with the activation of a two-component defense. In addition, they showed that α-hederin exhibits substantially higher biological activity than hederacoside C in causing hemolysis, the destruction of red blood cells through disruption of cell membranes due to its binding to cholesterol, a diagnostic property of saponins. α-Hederin is also more damaging than hederacoside C to plant cells. When the two components of the defense, hederacoside C and the β-glucosidase, were reconstituted in Nicotiana benthamiana, the α-hederin produced acted as a feeding deterrent to a generalist caterpillar. These results support the identification of hederacoside C as the protoxin and α-hederin as the biologically active derivative. However, the effectiveness of this system as a defense against the natural herbivores of common ivy remains to be determined.

On the biochemical side, Han et al. investigated the metabolism of these triterpenoid saponins. Careful gene expression analysis and heterologous expression in Escherichia coli revealed that three glucosyltransferases acting sequentially are responsible for the biosynthesis of the C-28 sugar chain, catalyzing the successive addition of glucose, glucose and rhamnose. Using activity-guided protein purification and proteomics, the authors also identified the β-glucosidase responsible for the activation of hederacoside C, which turned out unexpectedly to cleave the entire chain of three sugar molecules in a single reaction to form α-hederin. Glycosidases typically cleave individual sugar residues in a stepwise fashion, and this is also true of the glycosidases previously described in the activation of other two-component defenses. For example, activation of the cyanogenic diglycoside amygdalin to hydrogen cyanide requires the action of two different β-glucosidases in succession, each cleaving a single sugar residue (7). In common ivy, hydrolysis of all three C-28 sugar residues of hederacoside C in a single step may be critical in ensuring the rapid formation of the toxic product α-hederin.

This research also sheds light on the spatial separation between the two components of this activated defense system in the intact plant. Protein expression studies in N. benthamiana indicated that the enzymes responsible for the late steps of hederacoside C biosynthesis are localized in the cytosol or nucleus, whereas the protoxin is probably stored in the vacuole. The activating β-glucosidase, on the other hand, is localized in the chloroplast. This compartmentation likely prevents premature activation. Two-component defenses in different plant species have solved the separation problem in various ways at both the cellular or organellar levels. For example, a β-glucosidase hydrolyzing indolic glucosinolates in Arabidopsis thaliana is localized in the peroxisome of leaf epidermal cells (8), the glucosidase involved in hydrolyzing triterpenoid saponins in Medicago truncatula is localized in the nucleolus of root cell nuclei (4), and that hydrolyzing benzoxazinoids in maize leaves is located in the apoplast (9). In all these cases, the protoxin is likely stored in the vacuole to minimize the risk of hydrolysis prior to the major tissue damage caused by herbivore or pathogen attack.

A major unresolved question in the common ivy two-component defense system is why saponins are stored as glycosides containing an additional three-sugar side chain. In principle, a single sugar should be sufficient to produce an inactivated protoxin as observed in other two-component defense systems. The extended sugar side chains of triterpene saponins might increase the solubility or chemical stability of the protoxin, or decrease the risk of autotoxicity because it is more difficult to hydrolyze all three sugars accidentally at once. The three-sugar chain of glucose-glucose-rhamnose can be considered a security code that requires a very specific β-glucosidase as the key for cleavage. Such a system may also help counter the resistance of specialized herbivores that are able to metabolize the protoxins of two-component defenses to derivatives which cannot be activated by plant glycosidases (10, 11). Additional glycosylation may act against such herbivore counteradaptations, although more research on how herbivores metabolize triterpene saponins is required. Regardless of the reasons for having a three-sugar moiety to protect the toxin, once herbivore damage occurs, plants require rapid hydrolysis to activate the toxin. A β-glucosidase such as HhGH1, capable of cleaving the entire three-sugar chain in a single step, is ideally suited for this purpose (Fig. 1).

Fig. 1.

Activation of two-component defenses comparing the well-known glucosinolate-myrosinase system of the Brassicaceae with the newly described triterpenoid saponin system of the common ivy. In both systems, defensive compounds are stored as inactive precursors containing sugar groups that prevent autotoxicity. Upon tissue damage, specific glycosidases hydrolyze the sugar groups releasing toxic products that protect against herbivores. (A) In a canonical two-component defense system, such as the glucosinolate-myrosinase system, the glycosidase removes a single glucose residue, triggering the formation of toxic products. (B) In the triterpene saponins of the common ivy, an entire chain of three sugar residues is removed simultaneously by a specific glycosidase, thereby unmasking the toxin, which disrupts cell membranes.

From an evolutionary perspective, Han et al. found that the ability of the common ivy glycosidase HhGH1 to activate triterpenoid saponins was conserved among orthologs in other species of the family Araliaceae (the ginseng family) identified through genomic and synteny analyses. These plants also produce saponins like hederacoside C, suggesting a wider distribution of this two-component defense system. However, assays of these glycosidase orthologs revealed that some have diverged in catalytic activity and cleave sugars from the C-3 side chain of saponins rather than from the C-28 side chain. This finding suggests that there may be many distinct patterns for the synthesis and activation of two-component saponin defenses throughout the plant kingdom, making it even more important to understand the roles of different sugar chains in saponin activation and mode of action. Studies comparing herbivores resistant and susceptible to these defense systems may help explain the divergence in glycosylation and hydrolysis patterns. In considering the way plants have exploited triterpene saponins in defense, Han et al. highlight the long evolutionary history of triterpenes in plants, dating back to green algae, and the general ability of any triterpene structure with C-3 glycosylation and a hydrophobic backbone, such as α-hederin, to disrupt phospholipid bilayers. However, the use of saponins as activated defenses required the evolution of glycosylation at additional sites to render the molecules safe for storage, together with glycosidases capable of restoring toxicity. More detailed phylogenetic studies of the genes involved in protoxin biosynthesis and activation may provide greater clarity regarding the evolutionary timing of these events.

A new study by Han et al. on the triterpenoid saponins of the common ivy (Hedera helix) illuminates the two-component defense of this species in exquisite detail (5).

The study by Han et al. significantly expands the field of two-component defenses in terms of both the types of protoxins and the activating glycosidases employed in such systems, while also providing clues about their evolutionary origins. The authors’ success in reconstituting an activated triterpene saponin defense in a heterologous plant system such as N. benthamiana also suggests the feasibility of deploying two-component defenses for crop protection. Rather than relying on the external application of insecticides and fungicides, biosynthesis of toxins within the plant and their activation only after tissue damage places the active compounds precisely where they are needed while minimizing risks to nontarget organisms.