Jack C. Schultz

Each of the world’s 300,000 plant speciesis a target for attack from a range of nearly 400,000 species of plant-eating

insects. Herbivory is among the Earth’s mostimportant interactions in terms of number oftaxa, biomass and mass transfer, as well asevolutionary impact on plant traits, commu-nity structure and ecosystem function. Thepopulation biologists Paul Ehrlich and PeterRaven have even claimed that insect herbivory has generated much of terrestrialbiodiversity. Yet we are now realizing that the interactions between plants and insectsare far more dynamic than was previouslythought, and involve much shared chemistry.

Insects are responsible for 15% of theworld’s crop losses; even in natural systems,they consume 10% of plant production eachyear. Herbivory is limited directly or indirectlyby plant chemicals, which at first glance seemto have no role in normal plant metabol-ism but are highly active in animal tissues. Plant–herbivore interactions are oftendescribed as an ongoing biochemical warfarethat occurs on an evolutionary timescale.

But our perspective of plant defences andplant–insect interactions is now changing.Plants are not static, chemically defendedfortresses — they respond to attack withrapid, long-lasting, variable, and often specific biochemical, physiological and devel-opmental changes. Plants respond differen-tially to many stimuli, including variousinsect species. Every class of constitutivechemical defence responds to a physical insectattack, or to insect regurgitant or saliva, overperiods of minutes to days. The few ‘stealthy’insect species that fail to elicit any response atall are now considered exceptional.

Several signalling pathways coordinatethese responses. Fatty-acid signals (forexample, oxylipins, which are synthesized

from linolenic acid released from mem-branes by lipases) regulate expression ofdefence-related genes and are central to mostwound-mediated plant responses. Peptides,phenolics, terpenoids and classical planthormones (such as cytokinins and ethylene)also can help to coordinate plant responses.

As plants recognize pathogens by detect-ing specific molecules, it surely follows thatfeeding insects must likewise present identify-ing chemistry to plant receptors. But few suchresponse elicitors have been isolated frominsects. Salivary digestive enzymes have someactivity, as do several fatty-acid derivatives.The most complete study so far indicates thatphospholipid/amino acid conjugates foundin insect guts — and hence presumably in‘spit’ (or regurgitant) — elicit what may beherbivore-specific volatile emissions from thehost plant. These emissions in turn attractparasitoids, which kill the insect pests.

Plant-response elicitors, and the fatty-acid-based oxylipins used by plants them-selves to organize defence, are similar to signals used in animal defence systems.Eicosanoids in animals (including insects)match plant oxylipins in terms of biosynth-esis, structure, function and oxidative modification. These fatty-acid signals,including prostaglandins, leukotrienes andthromboxanes, mediate many immune andinflammation responses in animals.Eicosanoid receptors and their genes, as wellas their structure–function relationships,have been characterized in vertebrates. Muchof their activity is attributed to structural features that are shared by plant oxylipins.

Other chemical signals that are sharedacross the plant and animal kingdomsinclude peptides and glycoproteins, nitricoxide, reactive oxygen species, neuro-transmitters and plant growth hormones(auxins and cytokinins). Plant phenolics andsteroids interact with animal steroid recep-tors, and many plants produce moleculesthat act as insect hormones. The signallingactivities of many of these molecules areknown only in one of the kingdoms, but thisprobably reflects an absence of evidence —or study — rather than evidence of absence.

Plant defence responses require differen-tial gene expression in signalling, sensory andmetabolic pathways. It now seems that800–1,500 differentially regulated genes areinvolved in a plant’s response to a singleinsect, a complexity that matches that of animal responses.

Searching plant and animal genomes forsequences that are likely to share functionshas the potential to revise our view of plant defences completely— for example, asequence that encodes a putative glutamate

receptor has now been found in the genomeof the mustard weed (Arabidopsis thaliana).Ecologists have long known that plants produce neurotransmitters, including glu-tamate, that are presumably involved indefence against animals. The discovery ofsuch receptors in plants suggests that such‘secondary metabolites’ are more than purelydefensive, and that plants may be able to senseand respond to common animal signals.

There are important ecological and evolu-tionary implications of plants and animalssharing signalling systems. Theoretically, eachcould manipulate the other’s detection andresponse mechanisms. Insects present famil-iar signals to plants when they attack, andplant food contains chemicals that interferewith signalling in animals. This latter observa-tion is the basis of phytopharmacology, but isunappreciated in ecological and evolutionarycontexts. Although the phylogenetic distancebetween plants and herbivores is great, theorganisms share a common code-book, whichexplains how plants might identify insects,how both plants and insects might ‘jam’ eachother’s signalling, and how plants could wreakhavoc in insects by interfering with immuneresponses, reproduction and behaviour.

This new view of herbivory, emphasizingrapid, dynamic behaviours, shifts our focusfrom the chemical composition of plant tis-sues to the factors that elicit change and thesignals that coordinate it. Here there are asmany similarities between plants and insectsas there are differences, raising the possibilityof much ‘signal stealing’ and ‘biochemicalespionage’ between their two kingdoms. Perhaps, in terms of perception, signallingand biochemical behaviour, plants are reallyjust very slow animals after all. ■

Jack C. Schultz is in the Department of Entomology,Pennsylvania State University, University Park,Pennsylvania 16802, USA.

FURTHER READINGEhrlich, P. R. & Raven, P. H. Evolution 18,586–608 (1964).Karban, R. & Baldwin, I. T. Induced Responses toHerbivory (Univ. Chicago Press, Chicago, 1997).Shiu, S.-H. & Bleecker, A. B. Proc. Natl Acad. Sci. USA98, 10763–10768 (2001).Chiu, J., DeSalle, R., Lam, H. M., Meisel, L. & Coruzzi, G.Mol. Biol. Evol. 16, 826–838 (1999).

How plants fight dirtyconcepts

NATURE | VOL 416 | 21 MARCH 2002 | www.nature.com 267

BiochemicalecologyPlant–herbivore interactions arefrequently described as an ongoingbiochemical warfare that occurs onan evolutionary timescale.

A froghopper nymph coated in protective frothelicits a hugely complex defence by its plant host.

PAP

ILIO

/CO

RB

IS

© 2002 Macmillan Magazines Ltd