Hormone (Dis)harmony Moulds Plant Health and Disease

See allHide authors and affiliations

Science  08 May 2009:
Vol. 324, Issue 5928, pp. 750-752
DOI: 10.1126/science.1173771


Diseased plants often display phenotypes consistent with hormone perturbations. We review recent data that have revealed roles in plant-microbe interactions for cellular components and signaling molecules that previously were associated only with hormone signaling. A better understanding of cross-talk between hormonal and defense signaling pathways should reveal new potential targets for microbial effectors that attenuate host resistance mechanisms.

Plant pathogens can trigger a rich diversity of symptoms that indicate hormonal disorders, from galls and cankers of trees and shrubs, to “foolish seedling” disease of rice, and to premature senescence (or its converse, “green islands”) in many plants (Fig. 1). These hormonal aspects of pathogenesis were relatively neglected until recently, with the spotlight more on gene-specific interactions between host and pathogen. Integrated models now include activation of defense by pathogen-associated molecular patterns (PAMPs), suppression of defense by effectors, and recognition of particular effectors by specific resistance proteins (1, 2). Recent work is now revealing how pathogens enhance host susceptibility by manipulation of hormone signaling, and how hosts attenuate this manipulation. Here we assess current knowledge of hormonal signaling in plant-microbe interactions and highlight areas for future scrutiny, with a particular focus on the hormones jasmonate (JA), auxin, abscisic acid (ABA), and gibberellin (GA).

Fig. 1

Examples of plant growth distortions likely associated with disease-induced hormone perturbations. (A) Rust (Atelocauda digitata) on Acacia koa leaf, Hawaii. (B) Wheat infected with leaf rust (Puccinia striiformis) showing “green islands” in some interactions. (C) Wool sower gall maker (Callirhytis seminator) on oak. (D) Crown gall caused by Agrobacterium tumefaciens. (E) Witches broom in silver birch caused by Taphrina sp. (F) Gymnosporangium cornutum rust on European mountain-ash.

Antagonism of Salicyclic Acid Signaling as a Key Pathogen Virulence Mechanism

Salicylic acid (SA) and JA are signaling molecules in plant defense against biotrophic and necrotrophic pathogens, respectively (3). JA alone activates responses to wounding and herbivory, but in the presence of ethylene (Et), it activates defense against necrotrophs. JA signaling and SA signaling can compromise one another (4, 5). The result is that plant tissues usually can activate either SA or JA signaling, but not both.

Because the balance between JA and SA can determine whether the plant succumbs to infection, selection favors pathogens that can influence this balance in their favor. Pathogen modulation of other hormones, whether at the level of biosynthesis, bioactivity, or signaling, might also influence this balance. We propose that successful defense likely requires that the plant attenuate any microbe-induced hormonal perturbations.

Although the SA receptor remains elusive, the mechanism of JA, auxin, and GA signaling via targeted proteasome-mediated degradation through their respective Skp1-cullin-F box protein (SCF) E3 ubiquitin ligase complexes (6) is now better understood.

Jasmonates. Multiple JA derivatives exist, though knowledge of their bioactivity remains scarce. Many, but not all, jasmonate responses are mediated through the activity of the SCFCOI1 E3 ligase complex containing the coronatine-insensitive 1 (COI1) F-box protein (7). SCFCOI1 degrades a class of jasmonate signaling repressor (JAZ) proteins in the presence of jasmonoyl-l-isoleucine (JA-Ile) (8, 9). Many strains of the phytopathogenic bacterium Pseudomonas syringae (Pst) make coronatine (COR), a JA-Ile mimic. Activation of JA signaling by COR suppresses SA signaling, thus enhancing host susceptibility (10). COR binds with high affinity to a JAZ/SCFCOI1 complex, promoting degradation of JAZ proteins and thus providing a mechanistic link between COR action and virulence (11). The 12 Arabidopsis JAZ proteins form homo- and heterodimers, creating combinatorial diversity that could fine-tune responses during JA signaling (12). Additional receptors likely exist for JA relatives and other oxylipin signals because the JA-Ile–forming JAR1 mutant (13) fails to recapitulate all coi1 phenotypes and JA application induces coi1-independent JA responses (14, 15).

Auxin. Auxin promotes virulence during biotrophic interactions (1618). During defense triggered by flg22 (a peptide derived from the bacterial PAMP flagellin), mRNA expression of many auxin-signaling related genes is suppressed by the microRNA (miR393) that targets the auxin receptor F-box protein TIR1 and its paralogs AFB2 and AFB3. Plants that overexpress miR393 are more resistant to Pst, whereas plants that overexpress AFB1 (which is less sensitive to miR393 due to a nucleotide substitution) are more sensitive (17). PAMPs trigger SA accumulation (19), and SA can counteract auxin-induced virulence through generalized repression of auxin-related genes (including TIR1), which in turn stabilizes transcriptional repressors such as AXR2, thus inhibiting auxin responses (18).

GA. GA activates the SCFSLY1/GID2 complex to initiate degradation of the DELLA family of GRAS transcriptional repressors. A loss-of-function mutant lacking four of the five Arabidopsis DELLA-encoding genes exhibits enhanced resistance to Pst and, conversely, hypersusceptibility to the necrotrophic fungal pathogen Alternaria brassicicola. Infection by Pst results in elevated induction of the SA signaling marker PR1 and reduced and delayed expression of the JA/Et-induced PDF1.2 gene (20). DELLAs may potentiate JA signaling, suggesting that the rice “foolish seedling” disease pathogen Gibberella fujikoroi makes GA to eliminate DELLAs and compromise JA-mediated defense. Stress-induced DELLA accumulation increases the expression of genes encoding reactive oxygen species (ROS)–detoxification enzymes, thus reducing ROS levels (21). ROS may potentiate SA signaling, suggesting a mechanism whereby GA acts synergistically with SA by removing DELLAs, allowing more ROS accumulation because of reduced levels of ROS detoxification enzymes.

ABA. ABA influences many plant-pathogen interactions depending on pathogen life-style; Arabidopsis ABA-deficient mutants are hypersensitive to the oomycete Pythium irregulare (22) and fungus Leptosphaeria maculans (23) but more resistant to Botrytis cinerea (24). ABA positively regulates defense to Pst through regulation of pre-invasive stomata-based responses (25). However, at later stages, bacterial effectors activate ABA biosynthesis to overcome plant basal defenses (26). Botrytis and Cercospora species can themselves make ABA (27, 28). Because these pathogens are generally thought of as necrotrophs, this suggests a role for ABA during an early biotrophic phase before the pathogens switch to necrotrophy. ABA (and also methyl JA) can suppress callose deposition response to flg22 (26, 29).

A Central Role for DELLA Proteins?

An unexpected interaction between ABA, Et, and DELLA proteins has emerged. Both ABA and Et stabilize DELLA proteins (30, 31), which could potentiate JA signaling and attenuate SA signaling. Conceivably, Et switches JA signaling from wound responses to defense through its effect on DELLA stability. If DELLAs promote ROS detoxification and ABA stabilizes DELLAs, then ABA may promote susceptibility by reducing ROS levels, attenuating SA signaling. Auxin also promotes Et biosynthesis, and the resulting Et could stabilize DELLA proteins, thus promoting JA/Et signaling and attenuating SA signaling (32, 33). DELLA proteins up-regulate expression of XERICO, a putative E3 ligase that promotes ABA accumulation (34), thus establishing a potential positive-feedback loop to maintain ABA concentrations and ROS detoxification. ABA may also act at another node in the signaling network that promotes virulence. Pst infection of the aao3 mutant, which lacks the pathogen-inducible Arabidopsis aldehyde oxidase 3, leads to rapid accumulation of SA in the early stages of infection and failure to suppress PAMP-inducible genes (35). Notably, salt stress stabilizes DELLAs (36), and salt stress or exogenous ABA application suppresses chemically induced systemic immunity in Arabidopsis (37). In Fig. 2, we propose a model whereby DELLA proteins fine-tune the defenses mounted through the JA, Et, or SA pathways.

Fig. 2

Possible interactions between hormone signaling pathways are shown. JA/Et and SA signaling mutually interfere with each other. ABA and Et could strengthen JA/Et signaling and attenuate SA signaling via DELLA stabilization. Because auxin promotes Et biosynthesis, it might also interfere with SA signaling via DELLA stabilization. DELLAs increase expression of ROS detoxification mechanisms, attenuating redox stress and thus conceivably attenuating SA signaling.

Multiple Mechanisms of Control

In addition to these complicated signaling pathways, new insight is needed into other ways in which hormone and pathogen signaling intersect.

The role of conjugation. The Arabidopsis GH3 family of adenylate-forming enzymes, which includes JAR1, can adenylate a variety of plant hormones including JA, auxins, and SA (38). Inactivation of auxin via specific GH3s in Arabidopsis (39) or rice (40) conferred enhanced defense responses. However, GH3 variants also affect SA signaling (41, 42) or, in the case of GH3.5, both SA and auxin (39). PBS3, a GH3, acts not on SA but rather on 4-substituted benzoates, perhaps by priming or inducing SA biosynthesis (43). JAR1 conjugates isoleucine to JA, making active JA-Ile. Thus, GH3s can both activate and attenuate plant defense responses. SA can be conjugated to glucose and sequestered in the vacuole. The role of such conjugations in regulating signaling pathways is still not fully defined.

Regulation of de novosynthesis. We need more knowledge of hormone biosynthesis during the infection response. Although plant-pathogen interactions are usually studied in the leaf, many analyses of hormone biology are conducted on roots or seedlings, and the results might not be relevant to leaf processes. JA signaling can be modulated by SA that comes either from de novo synthesis or from large pools of glycosylated SA. Knowledge of how SA biosynthesis is regulated, in particular the signaling mechanism that activates isochorismate synthase (ICS1) (44), is still required.

Precursors for JA, SA, ABA, and GA are made in the chloroplast, then exported to the cytosol for the final synthetic steps. Infection with virulent Pst suppresses a range of nuclear-encoded chloroplast-associated genes (45, 46), though in contrast, stress-associated transcripts encoding components of ABA biosynthesis are up-regulated (46). Some bacterial type III effectors act in the chloroplast to suppress defense (47), indicating the importance of signaling between chloroplast and nucleus to modulate hormonal responses to pathogens.

The necessity for detailed temporal analysis. Exogenous hormone application reveals hormonal cross-talk on a gross scale but seldom mimics the temporal and spatial deployment of endogenous phytohormones. Exogenously applied ABA does not exert the same physiological effects as endogenous ABA (48). Pretreatment with GA enhanced susceptibility to Pst, suggesting GA may affect signaling of other hormones via destabilization of DELLAs (20). Similarly, when using mutants compromised in hormone biosynthesis or signaling, it is important to distinguish direct effects of a pathogen or defense modulation of specific hormones, from indirect effects arising from modification of the endogenous phytohormone balance.

We need studies that examine hormonal dynamics in detail through the course of an infection. Hormone endpoint measurements are inadequate. Microarray experiments on various pathogen infections over various time courses often suggest the involvement of hormone signaling, and temporally separated transient increases in hormone concentrations are likely to play an important role in configuring plant responses to microbial infection (49, 50). Challenges and opportunities remain in exploring the mechanisms underpinning the complex interactions between hormones and defense in whole plants.

References and Notes

  1. J.D.G.J. acknowledges support from the Gatsby Foundation and the Biotechnology and Biological Sciences Research Council (BBSRC), and helpful discussion with A.-R. Seilaniantz and R. Bari. M.R.G. acknowledges support from BBSRC.
View Abstract

Stay Connected to Science

Navigate This Article