Jasmonate signaling and manipulation by pathogens and insects

Introduction

Plants encounter various biotic and abiotic stresses throughout their life cycles. The lipid-derived hormones, jasmonates (JAs), enable plants to defend themselves against attacks by a wide variety of herbivores as well as necrotrophic pathogens that kill plant cells for nutrition (Howe and Jander, 2008; Antico et al., 2012; Erb et al., 2012; Campos et al., 2014; Yan and Xie, 2015). Insect herbivores affected by JA-induced plant defenses include leaf-eating insects, such as caterpillars and beetles as well as piercing-sucking insects, such as thrips, leafhoppers, spider mites, fungal gnats and mirid bugs, the stylet-mediated phloem-feeding aphids and whiteflies, as well as leafminers that feed on soft tissue between the upper and lower surfaces of leaves (Howe and Jander, 2008; Campos et al., 2014; Lu et al., 2015; Goossens et al., 2016). JA signaling also mediates plant defense against detritivorous crustaceans, vertebrate herbivores and molluscan herbivores (Farmer and Dubugnon, 2009; Mafli et al., 2012; Falk et al., 2014). In addition to herbivores, JA signaling mediates plant defense against necrotrophic pathogens, such as the bacterial pathogen Pectobacterium atrosepticum (syn. Erwinia carotovora subsp. atroseptica), fungal pathogens such as Alternaria brassicicola, Botrytis cinerea, Plectosphaerella cucumerina and Fusarium oxysporum, and oomycete Pythium spp (Campos et al., 2014; Yan and Xie, 2015). JA signaling has also been shown to mediate defense against some biotrophic and hemibiotrophic pathogens that obtain nutrients primarily from living plant cells; examples include rice resistance to Meloidogyne graminicola and Xanthomonas oryzae (Nahar et al., 2011; De Vleesschauwer et al., 2013). In addition to its role in regulating defense, JA is also required for plant reproduction and other growth and developmental processes, including lateral and adventitious root formation, seed germination, leaf senescence, and the formation of glandular trichomes, resin ducts, and nectaries (Wasternack and Hause, 2013; Campos et al., 2014; Kazan, 2015; Wasternack and Strnad, 2016). Interestingly, glandular trichomes, resin ducts, and nectaries can produce diverse compounds that are directly or indirectly involved in plant defense, linking the dual roles of JA in development and defense (Dicke and Baldwin, 2010; Campos et al., 2014).

A number of recent reviews have discussed topics ranging from JA biosynthesis to the molecular genetic dissection of JA signaling (Wasternack and Hause, 2013; Campos et al., 2014; Goossens et al., 2016). We refer readers to these excellent reviews. Here we will focus on recent literature on the elucidation of the structural mechanisms involved in transcriptional repression and activation of JA signaling, the various strategies used by pathogens and insects to manipulate JA signaling, and innovative approaches to interrupt pathogen hijacking of JA signaling for disease control.

Initiation of JA signaling during pathogen and herbivore attacks

It is now widely accepted that pathogen and herbivore attacks are associated with the generation of a variety of microbe/pathogen-associated molecular patterns (MAMPs/PAMPs) such as flagellin, herbivore-associated molecular patterns (HAMPs) such as insect secretions, and/or damage-associated molecular patterns (DAMPs) such as plant cell wall-derived oligogalacturonides and systemin or systemin-like peptides (Felton and Tumlinson, 2008; Mithöfer and Boland, 2008; Hogenhout and Bos, 2011; Yamaguchi and Huffaker, 2011; Campos et al., 2014; Heil and Land, 2014). These attacker-associated patterns are recognized by plant pattern recognition receptors (PRRs) located at the plant plasma membrane (Qi et al., 2006; Song et al., 2006; Yamaguchi et al., 2006; Brutus et al., 2010; Yamaguchi et al., 2010; Mousavi et al., 2013; Choi et al., 2014). Significant overlap of gene expression, including genes involved in defense hormone signaling, were observed across PAMP, HAMP and DAMP responses in several genome-wide transcriptome studies (Campos et al., 2014). Studies have also shown the rapid accumulation of JA in response to a wide range of MAMPs, HAMPs and DAMPs (Doares et al., 1995; McCloud and Baldwin, 1997; Li et al., 2002; Lee and Howe, 2003; Schmelz et al., 2003; Huffaker et al., 2006; Schmelz et al., 2007; Bonaventure et al., 2011; Yamaguchi and Huffaker, 2011; Huffaker et al., 2013; Campos et al., 2014; Kim et al., 2014). Plants can additionally sense the presence of insects via the pressure generated by an insect landing and walking, as well as mechanical wounding (Erb et al., 2012). In particular, mechanical wounding could cause rapid JA production and activation of JA signaling (Glauser et al., 2009; Koo et al., 2009; Acosta et al., 2013; Chauvin et al., 2013; Farmer et al., 2014).

The exact mechanism by which MAMP, HAMP and DAMP signaling leads to JA biosynthesis remains elusive. Several intracellular signals, including calcium ions, reactive oxygen species, mitogen-activated protein kinase (MPK) cascades and calcium-dependent protein kinases (CDPKs), have been implicated in the signal transduction from perception of these conserved patterns to induction of JA biosynthesis (Arimura and Maffei, 2010; Sato et al., 2010; Heinrich et al., 2011; Singh and Jwa, 2013; Romeis and Herde, 2014; Zebelo and Maffei, 2015; Ahmad et al., 2016). Nevertheless, it is not clear whether any of the enzymes involved in the biosynthesis of JA are regulated by CDPK/MPK-mediated phosphorylation, calcium/calmodulin binding or cellular redox changes, although JA-induced phosphorylation of JA signaling components have been observed (Katou et al., 2005; Zhai et al., 2013). JA is synthesized through the oxylipin biosynthesis pathway (Wasternack, 2007; Gfeller et al., 2010), starting with α-linolenic acid that is released from chloroplastic membranes following a pathogen or insect attack. Subsequent catalysis is processed by LIPOXYGENASE (LOX), ALLENE OXIDE SYNTHASE (AOS) and ALLENE OXIDE CYCLASE (AOC) to generate 12-oxo-phytodienoic acid (OPDA) in the chloroplast. OPDA is then transported into the peroxisome, where several cycles of β-oxidation take place and (+)-7-iso-JA is synthesized (Wasternack and Hause, 2013; Larrieu and Vernoux, 2016). After secretion into the cytosol, (+)-7-iso-JA is conjugated with the amino acid isoleucine (Ile) to generate JA-Ile, the most bioactive form of JA (Fonseca et al., 2009). Meanwhile, the JA cytosolic pool is converted into JA metabolites, for example through hydroxylation and/or carboxylation, in order to attenuate JA signaling, (Kitaoka et al., 2011; Koo et al., 2011; Heitz et al., 2012; Koo and Howe, 2012). A major area of future research would be to directly connect MAMP/HAMP/DAMP signaling, which appears to occur mainly in the plasma membrane, cytosol and nucleus, to JA biosynthesis, which occurs mainly in the chloroplast and peroxisome.

The core JA sensing and signaling module

Since the cloning of the CORONATINE INSENSITIVE1 (COI1) gene in 1998 (Xie et al., 1998), which was later found to encode a main component of the JA-Ile receptor complex (Thines et al., 2007; Katsir et al., 2008; Melotto et al., 2008; Sheard et al., 2010), the core elements of the JA signaling complex have been extensively characterized. This has led to a convincing framework for JA perception and initial signal transduction. In ‘stress-free’ plants with low levels of JA, JA-mediated responses are restrained by JASMONATE ZIM-domain (JAZ) proteins (Chini et al., 2007; Thines et al., 2007; Yan et al., 2007; Fonseca et al., 2009). JAZ proteins function as transcription repressors by binding and repressing MYCs, basic helix-loop-helix (bHLH) transcription factors (TFs) that belong to the IIIe subgroup of the bHLH family (Fig. 1A) (Heim et al., 2003; Boter et al., 2004; Lorenzo et al., 2004; Chini et al., 2007). The JAZ8 repressor can recruit the TOPLESS (TPL) family of corepressors directly through the TPL-binding ETHYLENE RESPONSE FACTOR (ERF)-ASSOCIATED AMPHIPHILIC REPRESSION (EAR) motif ‘LxLxL’ (Shyu et al., 2012). EAR motifs (‘LxLxL’ or ‘DLNxxP’) were also identified in JAZ5, JAZ6 and JAZ7 proteins (Kagale et al., 2010) and direct interaction between TPL and JAZ5/6 was detected in interactome experiments (Arabidopsis Interactome Mapping Consortium 2011; Causier et al., 2012). Most other JAZ proteins recruit TPL through binding, via the ZIM domain, to the EAR motif-containing adaptor protein NOVEL INTERACTOR OF JAZ (NINJA) (Kazan, 2006; Pauwels et al., 2010; Acosta et al., 2013). TPL proteins in turn recruit histone deacetylases (HDAs), such as HDA6 and HDA19, resulting in chromatin remodeling and suppression of JA-responsive gene expression (Zhou et al., 2005; Long et al., 2006; Wu et al., 2008). Additionally, JAZ1, JAZ3, and JAZ9 can directly interact with HDA6 leading to chromatin remodeling and repression of JA-responsive genes independently of NINJA and TPL proteins (Zhu et al., 2011).

In response to stress, plants synthesize JA-Ile, which directly promotes the interaction between JAZ and COI1, the F-box subunit of the SCF^COI1 ubiquitin E3 ligase (Fig. 1B) (Thines et al., 2007; Katsir et al., 2008; Melotto et al., 2008; Yan et al., 2009; Sheard et al., 2010). This hormone-dependent interaction leads to ubiquitination and degradation of JAZ proteins via the 26S proteasome, thereby derepressing MYC TFs (Chini et al., 2007; Thines et al., 2007; Yan et al., 2007). JA stimulates extensive transcriptional reprogramming through two branches of transcription activators. In one branch, JA signaling responds to wounding or herbivore attack and induces the production of defense proteins, such as VEGETATIVE STORAGE PROTEIN (VSP), via MYCs (Lorenzo et al., 2004). In the other branch, JA acts synergistically with ethylene (ET) upon necrotrophic pathogen attack and induces the production of defense proteins, such as PLANT DEFENSIN 1.2 (PDF1.2), via the APETALA2/ERF (AP2/ERF) TF family, such as ERF1 and OCTADECANOID-RESPONSIVE ARABIDOPSIS 59 (ORA59) (Pré et al., 2008; Zarei et al., 2011; Pieterse et al., 2012; Wasternack and Hause, 2013).

In addition to the COI1-JAZ-MYC canonical JA-mediated signaling cascade, JAZ repressors also interact with several other TFs, highlighting the multiple roles of JAZ proteins in regulating plant biology (Chini et al., 2016). The identification of these JAZ-binding TFs shed lights on how JA signaling modulates and integrates plant responses in response to developmental or environmental cues. Examples include the following: (i) Two members of the R2R3 MYB TF family, MYB21 and MYB24, interact with all JAZ proteins to regulate JA-mediated male fertility (Mandaokar et al., 2006; Song et al., 2011). (ii) Multiple JAZ proteins associate with WD-repeat/bHLH/MYB transcriptional complexes, which comprise WD-repeat TRANSPARENT TESTA GLABRA 1 (TTG1), IIIf bHLH TF TRANSPARENT TESTA 8 (TT8), GLABRA 3 (GL3) or ENHANCER OF GLABRA 3 (EGL3), and the MYB TF GLABRA 1 (GL1) or MYB75. This results in the repression of JA-mediated anthocyanin synthesis and trichome initiation (Traw and Bergelson, 2003; Pesch and Hulskamp, 2009; Qi et al., 2011b; Grebe, 2012; Qi et al., 2014). In this case, JAZ proteins directly interact with bHLH TFs and MYB TFs in the WD-repeat/bHLH/MYB complex and therefore interfere with the assembly and function of the WD-repeat/bHLH/MYB complex (Qi et al., 2011b). (iii) Four TFs from the IIId bHLH subfamily, bHLH3/JA-ASSOCIATED MYC2-LIKE 3 (JAM3), bHLH13/JAM2, bHLH14 and bHLH17/JAM1, which are phylogenetically closely related to MYC proteins, were found to interact with JAZ proteins. These bHLH TFs act as transcription repressors by antagonistically binding to the target sequence of MYC2 or the WD-repeat/bHLH/MYB complex and so negatively regulate JA-mediated responses (Nakata et al., 2013; Sasaki-Sekimoto et al., 2013; Song et al., 2013; Fonseca et al., 2014). (iv) Several JAZ proteins interact with the IIIb bHLH TFs, INDUCER OF CBF EXPRESSION 1 (ICE1) and ICE2, resulting in the repression of freezing tolerance in Arabidopsis (Hu et al., 2013). Binding specificities of these JAZ-bHLH complexes may be associated with fine tuning of JA-mediated responses (Chini et al., 2016). (v) In addition to the bHLH family, JAZ3 can interact with the YABBY (YAB) family TFs, FILAMENTOUS FLOWER (FIL)/YAB1 and YAB3, resulting in the repression of JA-mediated anthocyanin accumulation. Moreover, MYB75 is a direct transcriptional target of FIL, regulating anthocyanin accumulation (Boter et al., 2015). (vi) JAZ1, JAZ3, and JAZ9 can bind to ETHYLENE INSENSITIVE 3 (EIN3) and EIN3-LIKE 1 (EIL1) TFs that positively regulate the ET response, thereby suppressing the activity of EIN3 and EIL1 (Zhu et al., 2011). (vii) JAZ1, JAZ3, JAZ4, and JAZ9 proteins interact with the AP2 TFs, TARGET OF EAT 1 (TOE1) and TOE2. JA-triggered degradation of JAZ proteins release TOE1 and TOE2, both of which repress the transcription of FLOWERING LOCUS T (FT) and delay the flowering time of Arabidopsis (Zhai et al., 2015). (viii) WRKY57 is a repressor of JA-induced leaf senescence and interacts with JAZ4/8 or the AUXIN/IAA-INDUCIBLE (AUX/IAA) protein IAA29, regulating JA-auxin antagonism in leaf senescence (Jiang et al., 2014).

New structural insights into the COI1-JAZ-MYC signaling complex

While molecular, biochemical, and genetic studies support the view that COI1, JAZ, and MYC initiate a JA-dependent signaling cascade, the exact mechanisms of transcription repression or activation were until recently unclear. All JAZ proteins contain a conserved Jas motif at the C-terminus (Thines et al., 2007; Chini et al., 2007). Two studies elucidated the high-resolution structures of the COI1-Jas and Jas-MYC complexes (Sheard et al., 2010; Zhang et al., 2015a). Specifically, the crystal structure of the COI1-JA-Ile-Jas^JAZ1 complex shed light on how COI1 and JAZ proteins perceive JA-Ile (Sheard et al., 2010). In this structure, the three N-terminal α-helixes of COI1 bind to ASK1, a COI1-interacting subunit within the SCF^COI1 ubiquitin E3 ligase complex, whereas the 18 tandem leucine-rich-repeats (LRRs) at the C-terminus form a binding pocket for JA-Ile. On the other hand, the Jas^JAZ1 peptide adopts a bipartite structure in the presence of JA-Ile: (i) the five conserved N-terminal amino acids (‘ELPIA’) of the Jas^JAZ1 motif forms a loop that directly interacts with both JA-Ile and COI1 to trap JA-Ile into the ligand-binding pocket and (ii) the C-terminal region of the Jas^JAZ1 motif forms an α-helix for docking to the top surface of the COI1 LRR domain (Fig. 2A). Inside the ligand binding pocket, the amide and carboxyl groups of JA-Ile bind to three basic residues of COI1, R85, R348 and R409, via a salt bridge and hydrogen bond network (Fig. 2B). In addition to Y386, Y444 and R496, these arginine residues are critical for the COI1-Jas^JAZ1 interaction. Overall, the COI1-JAZ crystal structure is consistent with the radio-ligand binding assays showing both COI1 and JAZ proteins are required for high-affinity JA-Ile binding (Sheard et al., 2010). That is to say, JA-Ile is perceived as a high-affinity ligand by the COI1-JAZ coreceptor complex rather than COI1 or JAZ alone. In addition, the COI1-JA-Ile-JAZ structure provides a convincing explanation for the isomeric specificity of JA-Ile as a preferred ligand described previously (Staswick and Tiryaki, 2004; Fonseca et al., 2009). In particular, (3R,7S)-JA-Ile has a higher binding affinity than (3R,7R)-JA-Ile to the COI1-JAZ coreceptor. This is because the aliphatic chain from (3R,7R)-JA-Ile interferes with binding to COI1 and JAZ1. Analysis of the COI1-JA-Ile-Jas^JAZ1 crystal structure also led to the unexpected finding of inositol-1,2,4,5,6-pentakisphosphate (InsP₅) as a cofactor for the COI1-JAZ interaction. InsP₅ was found to interact with both the R206 residue of the Jas^JAZ1 peptide and three arginine residues of COI1, R85, R348 and R409, located at the bottom of the JA-Ile binding pocket (Fig. 2C). In addition to inositol phosphate, R206 of the Jas^JAZ1 peptide also directly interacts with the carboxyl group of JA-Ile. R206 cooperates with the three arginine residues of COI1, R85, R348 and R409, to form a salt bridge network that is required for the ligand-perception assembly. InsP binding specificity to the JA receptor is largely determined by the COI1 protein (Laha et al., 2015; Laha et al., 2016).

In addition to binding to COI1 to form the COI1-JAZ coreceptor that perceives JA-Ile, the Jas motif is also critical for interaction with MYC to mediate transcriptional repression (Cheng et al., 2011; Fernández-Calvo et al., 2011; Niu et al., 2011). How the same JAZ motif could engage COI1 for hormone-dependent activation of JA signaling and MYC for repression of JA signaling remained a mystery until a recent report with a detailed MYC3-Jas^JAZ1 complex structure (Zhang et al., 2015a). The N-terminus of MYC3 contains the JAZ-interacting domain (JID) and the transcription activation domain (TAD), which are required and sufficient for the JAZ-MYC interaction (Fernández-Calvo et al., 2011; Çevik et al., 2012; Chen et al., 2012). Zhang and colleagues (2015a) solved the structures of the Jas^JAZ1-MYC3 and Jas^JAZ9-MYC3 complexes, revealing that the N-terminus of MYC3 forms a helix-sheet-helix sandwich fold, in which the central five-stranded anti-parallel sheets are surrounded by eight helices (Fig. 2D). The α4 helix of TAD forms a groove with JID. In the MYC3-Jas^JAZ1 and MYC3-Jas^JAZ9 complex structures, the Jas peptide forms a single, continuous helix to occupy in the groove formed by JID and TAD and becomes an integral part of the MYC3 N-terminal fold. Correspondingly, MYC3 undergoes a dramatic conformational change upon engaging the Jas motif sequence (Zhang et al., 2015a).

The JAZ-MYC3 structure provides mechanistic insights into the transcriptional repression and hormone-dependent activation of the JA pathway. First, comparison of the structures of the COI1-JA-Ile-JAZ and JAZ-MYC complexes reveals extensive overlap in COI1 and MYC binding to the Jas motif. However, the COI1-JAZ interaction involves more amino acid residues in the Jas domain compared with the Jas-MYC interaction. Specifically, the N-terminal portion (‘ELPIA’ in JAZ9) of the Jas motif that is critical for JAZ binding to JA-Ile/COI1 is not essential for the MYC-JAZ interaction. Furthermore, in contrast to the continuous helix conformation of the Jas motif in the JAZ-MYC complex, the Jas motif in the COI1-JAZ complex adopts a bipartite conformation (Fig. 2A), involving JA-Ile-dependent unwinding of the N-terminal portion of the Jas helix. It is likely that the more extensive interactions in the COI1-JA-Ile-JAZ complex allow COI1 to compete with MYC for JAZ interaction upon JA-Ile stimulation. Secondly, the Jas helix occupies the groove formed by JID and TAD in the JAZ-MYC3 structure and makes direct contact not only with JID but also with TAD, which is required for transcriptional activation. This unexpected finding suggests the possibility of direct competition between the JAZ repressor and transcription coactivators. Indeed, Zhang and colleagues showed that JAZ could directly inhibit the interaction between MYC3 and MED25 (Zhang et al., 2015a), which is a component of the Mediator co-activator complex required for JA gene expression (Çevik et al., 2012; Chen et al., 2012). Thus, in addition to JAZ-mediated recruitment of TPL/NINJA corepressors/adaptors for chromatin-based transcriptional repression, JAZ repressors also directly inhibit interactions between MYC proteins and MED25. Such a dual transcriptional repression mechanism may be the key to ensuring tight and dynamic control of JA responses.

Pathogen and insect manipulation of JA biosynthesis and signaling

Perspectives

Clearly, manipulation of SA-JA antagonism has emerged as a major theme in plant interactions with pathogens, insects and nematodes. The core components of JA signaling, particularly COI1 and JAZ coreceptor proteins, appear to be common targets of virulence factors from biotrophic and hemibiotrophic pathogens. Conventional activation of either the JA or SA signaling pathway, through genetic or chemical manipulation, encounters a risk of improving plant defense against one attacker, but inherently priming plant susceptibility to other attackers, illustrating the complexity and vulnerability of the plant defense network. For example, classical loss-of-function coi1 mutants exhibit high level resistance to P. syringae but are greatly compromised in defense against chewing insects and necrotrophic pathogens (Stintzi et al., 2001; Glazebrook, 2005; Robert-Seilaniantz et al., 2011; Thaler et al., 2012).

How do we solve this dilemma? One approach would be to modify specific JA signaling components to be insensitive to manipulation by pathogen or insect virulence factors but preserving the functions of these signaling components in the perception and signal transduction of endogenous JA. A recent study illustrates the feasibility of this approach. Guided by the crystal structure of the COI1-JAZ coreceptor and evolutionary clues from the putative ligand-binding pockets of COI1 paralogs of low plant species, Zhang et al. (2015b) were able to make a single nucleotide/amino acid substitution, A384V, in the ligand-binding pocket of the Arabidopsis COI1 protein that allows for sufficient signal transduction of endogenous JA-Ile, but has greatly reduced sensitivity to the P. syringae toxin COR (Zhang et al., 2015b). Consequently, transgenic Arabidopsis plants expressing the engineered COI1^A384V receptor not only maintained male fertility and a high level of insect defense but also gained resistance to the hemibiotrophic COR-producing pathogens P. syringae pv. tomato DC3000 and P. syringae pv. maculicola ES4326. This result provides a proof-of-concept demonstration for modifying the host targets of pathogen virulence factors as a promising new approach to broaden the capacity of host defense against highly evolved pathogens.

As mentioned above, HopZ1a from Psy A2 and HopX1 from Pta 11528 promote the degradation of JAZ proteins through direct interaction, resulting in activation of JA signaling (Jiang et al., 2013; Gimenez-Ibanez et al., 2014). MiSSP7 from Laccaria bicolor interacts with the Populus PtJAZ6 protein, inhibits JA-induced degradation of PtJAZ6 and promotes symbiosis (Plett et al., 2014). In these cases, innovative JAZ-based methods could be developed to counter pathogen virulence and enhance beneficial symbiosis. Overall, growing knowledge of JA signaling and pathogen hijacking has begun to reveal disease vulnerable targets that may be repaired as a novel strategy for defense reinforcement.