1 Introduction

Secondary metabolites play diverse roles in plants. Flowers for example synthesize and accumulate anthocyanin pigments to attract pollinating insects. The biosynthesis of anthocyanins is under tight developmental control (Grotewold, 2006). Other compounds, including alkaloids in many plant species (Facchini, 2001) and glucosinolates in the Brassicaceae (Halkier and Gershenzon, 2006), function in protection against microorganisms or herbivores.

Biosynthesis of defensive secondary metabolites is often induced in plants following attack by microorganisms and/or herbivores. Primary signals specifying attack by fungal or bacterial microorganisms which are recognized by the plant are called elicitors, pathogen-associated molecular patterns (PAMPs) or microbe-associated molecular patterns (MAMPs) (Jones and Dangl, 2006). MAMPS that elicit a defence response include certain molecules derived from the microbial cell wall or the bacterial flagella. Attack by herbivores is recognized among others by certain molecules released from damaged plant cell walls and by certain molecules present in insect oral secretions. Elicitors activate signal transduction pathways that generate secondary signals within plants (Zhao et al., 2005). Three major plant secondary signalling molecules are jasmonates (JAs; Turner et al., 2002; Wasternack, 2007; Balbi and Devoto, 2008), ethylene (Wang et al., 2002) and salicylic acid (Shah, 2003). Production of these hormones leads to cascades of events responsible for the physiological adaptation of the plant to the external stress. The JAs, ethylene and salicylic acid signal transduction pathways act synergistically or antagonistically in a variety of responses, leading to fine-tuning of the complex defence response (Kunkel and Brooks, 2002).

Among these three key regulatory signals, by far the most important molecules for induction of secondary metabolism are the JAs. JAs have been found to induce the biosynthesis of a variety of secondary metabolites in different plant species, including alkaloids, terpenoids, glucosinolates and phenylpropanoids (Memelink et al., 2001; Zhao et al., 2005). This chapter will review recent advances in our understanding of the mechanism of action of JAs in induction of alkaloid metabolism in tobacco and in Catharanthus roseus (Madagascar periwinkle). These advances have been enabled by pioneering research on signal transduction of JAs using the model plant Arabidopsis thaliana (Katsir et al., 2008a).

2 Jasmonates Are Essential for Elicitor Signal Transduction

Addition of elicitors is a common method of enhancing secondary metabolism in plant cell cultures for metabolic, enzymatic or regulatory studies. In its broadest definition, an elicitor is any compound or mixture of compounds that induces a plant defence reaction. Most elicitors used in plant research originate from microorganisms but others are derived from the plant cell wall. In addition, a variety of abiotic elicitors has been used, such as heavy metals.

Elicitors are often applied in the form of crude mixtures, such as a fungal cell wall extract. In a few cases, elicitors have been purified to homogeneity. In alkaloid research, an extract from baker’s yeast is commonly used as an elicitor. Yeast extract contains several components that can elicit plant defence responses, including chitin, N-acetylglucosamine oligomers, β-glucan, glycopeptides and ergosterol. In addition, a low molecular weight component, which is probably a small peptide, induces the expression of terpenoid indole alkaloid biosynthesis genes in cells of C. roseus (Menke et al., 1999b).

Biotic elicitors induce a defence reaction in plant cells because they are recognized as “non-self” in the case of microbial elicitors or as “abnormally modified self” in the case of plant cell wall fragments. Intensive research efforts, including pharmacological studies, have uncovered components of the signal transduction pathway connecting elicitor perception to induction of defence genes (Zhao et al., 2005). In several elicitor responses, including secondary metabolite production (methyl)jasmonic acid and some of its bioactive precursors and derivatives play key roles as intermediate signals. In different plant species, elicitors were shown to induce accumulation of endogenous jasmonic acid, and (Me)jasmonic acid itself increased secondary metabolite production (Menke et al., 1999b; Zhao et al., 2005). In addition, in several species it was shown that blocking jasmonate biosynthesis abol-ished elicitor-induced metabolite accumulation and the expression of biosynthesis genes (Menke et al., 1999b; Zhao et al., 2005). Elicitors induce many intracellular events, including an increase in cytoplasmic calcium concentration, ion transport, production of reactive oxygen species and protein phosphorylation. How these events are exactly coupled to induced jasmonate biosynthesis remains largely unknown. A detailed review of intracellular events triggered by elicitors and their possible role in signal transduction is presented by Zhao et al. (2005).

The control points that govern the synthesis and accumulation of JAs remain to be identified. Timing and control of jasmonate biosynthesis suggest several ways in which jasmonate signalling might be modulated during stress perception. One level of control in jasmonate biosynthesis and/or signalling might be the sequestration of biosynthetic enzymes and substrates inside the chloroplasts (Stenzel et al., 2003). In this way, jasmonate biosynthesis and signalling would only be activated by the availability of substrate upon cellular decompartmentalization during wounding or pathogen attack. However, wounding induces the expression of several jasmonate biosynthesis genes (Turner et al., 2002), suggesting that, at least partly, the wound-induced production of JAs is a result of the increased transcription of genes encoding the jasmonate biosynthesis pathway enzymes and their subsequent de novo protein synthesis. In addition, JAs themselves induce the expression of genes involved in jasmonate biosynthesis (Turner et al., 2002), indicating the existence of a positive feedback regulatory mechanism for jasmonate biosynthesis in which JAs stimulate their own production.

3 Jasmonate Biosynthesis

JAs, including jasmonic acid (JA) and several of its cyclic precursors and derivatives, constitute a family of bioactive oxylipins that regulate plant responses to environmental and developmental cues (Turner et al., 2002; Wasternack, 2007; Balbi and Devoto, 2008). These signalling molecules affect a variety of developmental processes including fruit ripening, production of viable pollen, root elongation, and tendril coiling. In addition and more importantly for this review, JAs regulate responses to wounding and abiotic stresses, and defence against insects and necrotrophic pathogens.

An important defence response is the induction of secondary metabolite accumulation, which depends on JAs as a regulatory signal. JAs are fatty acid derivatives which are synthesized via the octadecanoid pathway (Fig. 1). Most of the enzymes of this pathway leading to jasmonate bio-synthesis have been identified by a combination of biochemical and genetic approaches (Wasternack, 2007). The enzymes leading to JA biosynthesis are located in two different subcellular compartments. The octadecanoid pathway starts in the chloroplasts with phospholipase-mediated release of α-linolenic acid from membrane lipids. The fatty acid α-linolenic acid is then converted to 12-oxo-phytodienoic acid (OPDA) by the sequential action of the plastid enzymes lipoxygenase (LOX), allene oxide synthase (AOS), and allene oxide cyclase (AOC). The second part of the pathway takes place in the peroxisomes. OPDA is transported from the chloroplasts to the peroxisomes where it is reduced by OPDA reductase (OPR3) to give 3-oxo-2(2’[Z]-pentenyl)-cyclopentane-1-octanoic acid (OPC:8), followed by three rounds of beta-oxidation involving three enzymes to yield (+)-7-iso-jasmonic acid which equilibrates to the more stable (-)-JA.

Fig. 1
figure 1_8

Schematic representation of the octadecanoid pathway leading to jasmonic acid biosynthesis. 12-OH-JA, 12-hydroxy-jasmonic acid; AOC, allene oxide cyclase; AOS, allene oxide synthase; JA, jasmonic acid; JAR1, enzyme responsible for the conjugation of JA with isoleucine (JA-Ile); JMT, S-adenosyl-l-methionine:jasmonic acid carboxyl methyl transferase; LA, α-linolenic acid; LOX, lipoxygenase; MeJA, methyl jasmonate; OPDA, 12-oxo-phytodienoic acid; OPR3, OPDA reductase3; PL, phospholipase

Subsequently, JA can be metabolized in the cytoplasm by at least seven different reactions. Well-characterized reactions include methylation to methyl-jasmonate (MeJA) by S-adenosyl-l-methionine:jasmonic acid carboxyl methyl transferase (JMT), conjugation to amino acids by JA amino acid synthase (JAR1) or hydroxylation to 12-hydroxyjasmonic acid (12-OH-JA). OPDA, JA, MeJA and JA-Ile are active signalling molecules, whereas 12-OH-JA is thought to be a biologically inactive derivative (Wasternack, 2007).

4 Jasmonate Perception and Signalling in Arabidopsis and Tomato

How JAs induce gene expression has been mainly unravelled in studies using Arabidopsis and tomato (Solanum lycopersicum) (Katsir et al., 2008a). To identify molecular components of jasmonate signal transduction, screenings for Arabidopsis mutants that are insensitive to (Me)JA or to coronatine (a bacterial toxin which is a structural and functional analogue of JA-Ile) or that show constitutive jasmonate responses have been performed (Lorenzo and Solano, 2005). A number of mutants have been characterized, and will be discussed below.

The coronatine insensitive1 (coi1) mutant was isolated in a screen for Arabidopsis mutants insensitive to root growth inhibition by coronatine (Feys et al., 1994). The coi1 mutant is also insensitive to JAs (Feys et al., 1994), is defective in resistance to certain insects and pathogens and fails to express jasmonate-regulated genes (Turner et al., 2002). The COI1 gene encodes an F-box protein (Xie et al., 1998). F-box proteins associate with cullin, Skp1 and Rbx1 proteins to form an E3 ubiquitin ligase known as the SCF complex, where the F-box subunit functions as the specificity determinant targeting proteins for ubiquitin-mediated proteolysis by the 26S proteasome (del Pozo and Estelle, 2000). Co-immunoprecipitation experiments showed that COI1 associates in vivo with Skp1, cullin and Rbx1 proteins to form the SCFCOI1 complex (Devoto et al., 2002; Xu et al., 2002). Therefore, the requirement for COI1 in jasmonate-dependent responses indicates that ubiquitin-mediated protein degradation is a crucial event in jasmonate signalling. Plants that are deficient in other components or regulators of SCF complexes, including AXR1, COP9 and SGT1b, also show impaired jasmonate responses (Lorenzo and Solano, 2005). The existence of a COI1 function that is conserved in species other than Arabidopsis was demonstrated by the identification of COI1 homologues in tomato (Li et al., 2004), tobacco (Shoji et al., 2008) and Nicotiana attenuata (Paschold et al., 2007). COI1 is a component that is specific to the JA pathway, whereas SGT1b and AXR1 are shared by other signalling pathways. Mutations in AXR1 or SGT1b have pleiotropic effects that impair plant responses not only to JA but also to auxin and pathogens, suggesting that both SGT1b and AXR1 are regulators of SCF complexes and are involved in several different signalling pathways (Austin et al., 2002; Azevedo et al., 2002; Gray et al., 2003).

A particularly effective screen for jasmonate signalling mutants has been described by Lorenzo et al. (2004). Screening for mutants affected in JA-induced root growth inhibition in an ethylene-insensitive3 (ein3) background resulted in the identification of five loci called JA-insensitive (JAI) 1–5. The JAI1 locus corresponds to the AtMYC2 gene (Lorenzo et al., 2004), encoding a basic-Helix-Loop-Helix (bHLH) transcription factor which regulates a subset of jasmonate-responsive genes involved in wounding responses. The JAI2 locus corresponds to the previously characterized JAR1 gene (Staswick et al., 1992), encoding an enzyme that couples JA to amino acids with a preference for isoleucine (Staswick and Tiryaki, 2004). The JAI4 locus corresponds to the SGT1b gene (Lorenzo and Solano, 2005). The JAI5 locus corresponds to the COI1 gene (Lorenzo et al., 2004).

Recently, the gene affected in the jai3 mutant was identified. It encodes a protein with a zinc finger-like ZIM motif (Chini et al., 2007). There are several related genes in Arabidopsis forming a gene family called ZIM or TIFY (Vanholme et al., 2007). The members that are induced at the gene expression level by JAs are called Jasmonate ZIM domain (JAZ) proteins (Chini et al., 2007; Thines et al., 2007). They contain in addition to the highly conserved central ZIM motif a highly conserved C-terminal Jas motif and a less conserved N-terminal region. In the jai3 mutant an aberrant protein is expressed with a deletion of the C-terminal domain including the Jas motif. The wild-type JAI3 (or JAZ3) protein is rapidly degraded in response to JA in a COI1-dependent manner, whereas the jai3 mutant protein is stable. The JAI3 protein was shown to interact in vitro and in yeast with AtMYC2. Based on these findings it was postulated that JAI3 is a repressor of AtMYC2 which is rapidly degraded in response to JA thereby activating AtMYC2 (Fig. 2; Chini et al., 2007).

Fig. 2
figure 2_8

Model for jasmonate signal transduction leading to expression of AtMYC2-regulated genes. Although depicted as a single protein, COI1 forms part of the E3 ubiquitin ligase SCFCOI1.. In the absence of JA-Ile, JAZ repressors interact with AtMYC2 maintaining this transcription factor inactive. In the presence of JA-Ile, the F-box protein COI1 binds to JAZ proteins, which results in their ubiquitination by the SCFCOI1 complex and their degradation by the 26S proteasome. AtMYC2 is liberated and activates transcription of target genes, including genes encoding JAZ proteins, resulting in a negative feedback loop

In independent studies, members of the JAZ gene family in Arabidopsis were characterized as being predominant among genes induced in anthers after 30 min of JA treatment (Mandaokar et al., 2006). Subsequent study of the family member JAZ1 demonstrated that it is rapidly degraded in response to JA in a COI1-dependent manner (Thines et al., 2007). On the other hand a deletion derivative of JAZ1 lacking the C-terminal domain is stable.

Interestingly, these authors were able to detect interaction between JAZ1 and COI1 in a yeast two-hybrid assay in the presence of JA conjugated to Ile (JA-Ile) in the yeast growth medium or in an in vitro pull-down assay in the presence of JA-Ile. No interaction was detected in the presence of OPDA, JA, MeJA or JA conjugated to Trp or Phe, whereas JA-Leu was about 50-fold less effective in promoting interaction between COI1 and JAZ1 than JA-Ile. JA-Ile and JA-Leu are products of the JAR1-mediated conjugation reaction (Staswick and Tiryaki, 2004). JA-Ile and coronatine also promote the interaction between JAZ3 and JAZ9 in a yeast two-hybrid assay, whereas JA or MeJA are ineffective (Melotto et al., 2008).

Using tomato SlCOI1 and SlJAZ1, it was shown that the complex binds radiolabeled coronatine (Katsir et al., 2008b). Binding can be displaced with unlabeled coronatine or JA-Ile. These experiments show that COI1 is the receptor for at least certain JAs including JA-Ile, as well as for the microbial JA-Ile mimic coronatine. The C-terminal domains containing the conserved Jas motif of tomato JAZ1 (Katsir et al., 2008b) and Arabidopsis JAZ1, JAZ3 and JAZ9 (Melotto et al, 2008) were shown to be necessary and sufficient for binding to COI1 in a JA-Ile or coronatine-dependent manner.

The expression of the JAZ genes in Arabidopsis is induced by JA (Mandaokar et al., 2006; Chini et al., 2007; Thines et al., 2007) and is controlled by AtMYC2 (Chini et al., 2007). AtMYC2 and JAZ proteins therefore form a jasmonate-responsive oscillator, where JAZ proteins negatively regulate AtMYC2 activity at the protein level, JAs cause JAZ degradation and AtMYC2 activation, and AtMYC2 switches on the expression of JAZ repressors at the gene level (Fig. 2).

The picture that emerges for jasmonate signal transduction is highly reminiscent of auxin signal transduction, which involves auxin-responsive degradation of AUX/IAA repressor proteins via the F-box protein TIR1 (Guilfoyle, 2007). TIR1 is the auxin receptor (Kepinski and Leyser, 2005; Dharmasiri et al., 2005) with auxin acting as the molecular glue between TIR1 and AUX/IAA proteins (Tan et al., 2007). COI1 is the closest relative to TIR1 that is not related to auxin perception among the about 700 members of the Arabidopsis F-box protein family (Gagne et al., 2002). JA-Ile enhances the interaction between COI1 and JAZ1, JAZ3, JAZ9 and possibly other JAZ family members (Fig. 2). It has been proposed that different biologically active JAs could promote the binding between COI1 and specific JAZ family members, and that these family members could act as repressors of specific downstream targets, presumably other transcription factors (Thines et al., 2007).

Challenges are to determine whether different JAs can indeed promote interaction of COI1 with specific JAZ family members, and to find out what are the specific targets of each member of the JAZ family of repressors. It is also conceivable that JA-Ile or other biologically active JAs enhance binding between COI1 and hitherto unidentified repressors distinct from the JAZ proteins.

5 Jasmonate Signalling in Terpenoid Indole Alkaloid Biosynthesis in Catharanthus roseus

Catharanthus roseus has the genetic potential to synthesize over a hundred terpenoid indole alkaloids (TIAs; van der Heijden et al., 2004). Several of the terpenoids that are known to be produced by C. roseus have pharmaceutical applications, including the monomeric alkaloids serpentine and ajmalicine, which are used as a tranquilizer and to reduce hypertension, respectively, and the dimeric alkaloids vincristine and vinblastine, which are potent antitumour drugs. TIAs consist of an indole moiety derived from tryptophan, and a terpenoid group derived from geraniol (Fig. 3). The condensation of the tryptophan derivative tryptamine with the terpenoid derivative secologanin is performed by the enzyme strictosidine synthase (STR) and results in the synthesis of 3α(S)-strictosidine. TIA biosynthesis has been shown to be induced by MeJA in developing seedlings and in cell cultures (Memelink and Gantet, 2007).

Fig. 3
figure 3_8

Biosynthesis pathway for terpenoid indole alkaloids in C. roseus. Unbroken arrows indicate single enzymatic conversions and broken arrows indicate multiple enzymatic conversions. The structures of ajmalicine and vinblastine are shown. STR, strictosidine synthase

Over ten genes have now been cloned from the estimated total number of around 25 genes encoding enzymes involved in TIA biosynthesis (Memelink and Gantet, 2007). In addition, genes acting in primary precursor pathways leading to the formation of tryptophan (van der Fits and Memelink, 2000) and geraniol (Chahed et al., 2000; Veau et al., 2000) have been cloned. All TIA biosynthesis genes tested are induced by MeJA in Catharanthus suspension-cultured cells (van der Fits and Memelink, 2000). In addition, MeJA induces genes in primary metabolism leading to the formation of TIA precursors. This presents a good example of the profound effect of JAs on plant metabolism at the level of gene expression.

The promoter of the STR gene has been studied in detail to identify elicitor- and jasmonate-responsive sequences. A region close to the TATA box called jasmonate- and elicitor-responsive element (JERE) was found to dictate elicitor- and jasmonate-responsive reporter gene activation (Menke et al., 1999a). The JERE interacts with two transcription factors called Octadecanoid derivative-Responsive Catharanthus AP2-domain proteins (ORCAs). ORCA2 was isolated by yeast one-hybrid screening of a Catharanthus cDNA library with the JERE as bait (Menke et al., 1999a) and ORCA3 was isolated by a genetic T-DNA activation tagging approach (van der Fits and Memelink, 2000). Both belong to the AP2/ERF family of transcription factors, which are unique to plants and are characterized by the APETALA2/Ethylene Response Factor (AP2/ERF) DNA-binding domain. In transient assays both ORCA proteins transactivate STR promoter activity via specific binding to the JERE. Over-expression of ORCA3 (van der Fits and Memelink, 2000) or ORCA2 (unpublished results) in stably transformed Catharanthus cells leads to elevated expression levels of STR and several other TIA biosynthesis genes.

Importantly, the expression of the ORCA2 and ORCA3 genes themselves is induced by JAs (Menke et al., 1999a; van der Fits and Memelink, 2000, 2001). This suggests that JAs induce alkaloid metabolism by increasing the amount of the ORCA regulatory proteins. To study how the ORCA3 gene is regulated, its promoter was used in loss- and gain-of- function experiments to identify a 74 bp D region containing a jasmonate-responsive element (JRE; Vom Endt et al., 2007). The JRE is composed of two important sequences, a quantitative sequence responsible for a high level of expression, and a qualitative sequence that acts as an on/off switch in response to MeJA. Using the JRE in yeast one-hybrid screening of Catharanthus cDNA libraries, several proteins belonging to the AT-hook class of DNA-binding proteins were isolated, which were found to interact specifically with the quantitative sequence within the JRE (Vom Endt et al., 2007).

The qualitative element consists of a G-box-like sequence (AACGTG). In a yeast one-hybrid screen of a Catharanthus cDNA library using a G-box as bait several genes encoding transcription factors were isolated, including five members of the bHLH family (Pré et al., 2000). One of these proteins called CrMYC2 is a close homologue of AtMYC2, and the corresponding gene was the only one among those five genes that was rapidly and strongly induced by MeJA in Catharanthus cells (unpublished results). CrMYC2 bound specifically to the qualitative sequence in the JRE of the ORCA3 promoter in vitro, and it transactivated reporter gene expression specifically via interaction with the qualitative sequence in transient assays in Catharanthus cells (unpublished results). Silencing the expression of CrMYC2 via RNAi strongly reduced the MeJA-responsive expression of the ORCA2 and ORCA3 genes in stably transformed Catharanthus cells (unpublished results). This demonstrates that MeJA-responsive ORCA expression is controlled by CrMYC2 (Fig. 4).

Fig. 4
figure 4_8

Model for jasmonate signal transduction leading to the expression of terpenoid indole alkaloid biosynthesis genes in Catharanthus roseus. As depicted in Fig. 1, a bioactive jasmonate species enhances the interaction between CrCOI1 and CrJAZ, leading to degradation of the latter proteins. CrMYC2 then activates transcription of the genes encoding the AP2/ERF-domain transcription factors ORCA2 and ORCA3, which in turn activate the expression of terpenoid indole alkaloid biosynthesis genes. CrMYC2 also activates transcription of CrJAZ genes as part of a negative feedback loop. The position of CrCOI1 in this signal transduction pathway is hypothetical as indicated by the question mark. Solid lines indicate interactions between proteins and broken lines indicate interactions between proteins and genes

Transcript profiling of Catharanthus cells treated with MeJA identified two members of the Catharanthus JAZ family (Rischer et al., 2006). The corresponding genes were rapidly and strongly induced by MeJA in wildtype Catharan­thus cells (Rischer et al., 2006; unpublished results). This response was strongly reduced in the transgenic cell line with silenced CrMYC2 expression, showing that MeJA-responsive CrJAZ gene expression is regulated by CrMYC2 (unpublished results). A CrJAZ1-GFP fusion protein was shown to be rapidly degraded in Catharanthus cells treated with MeJA, whereas a deletion derivative lacking the C-terminal Jas domain was much more stable. In transient reporter gene assays, the CrJAZ proteins abolished the activating activity of CrMYC2. This shows that the full-length CrJAZ proteins are indeed repressors that negatively regulate CrMYC2 activity. Yeast two-hybrid assays showed that both CrJAZ proteins were able to interact with CrMYC2 (unpublished results).

Our results prompted us to propose the model shown in Fig. 4 for signalling by JAs in C. roseus leading to alkaloid biosynthesis. Perception of a bioactive jasmonate derivative by CrCOI1 results in the degradation of CrJAZ proteins. CrMYC2 then activates the expression of the ORCA genes, which in turn activate the expression of a subset of TIA biosynthesis genes. Simultaneous activation of JAZ genes by CrMYC2 restores the un-induced situation by inhibition of CrMYC2 activity. The identity of the active jasmonate signalling molecule(s) and the involvement of CrCOI1 in this sequence of events still need to be experimentally confirmed.

6 Jasmonate Signalling in Tobacco Alkaloid Biosynthesis

The main alkaloid found in tobacco plants, nicotine, is composed of a pyrrolidine ring and a pyridine ring (Fig. 5). The pyrrolidine moiety is derived from N-methylputrescine, which is formed from putrescine by putrescine N-methyltransferase (PMT) (Katoh et al., 2005). The pyridine moiety of nicotine is derived from nicotinic acid. Nicotine is exclusively synthesized in the roots and is translocated to the leaves via the xylem. Multiple structural genes for nicotine biosynthesis enzymes, including PMT, are transcriptionally activated by exogenous application of jasmonates in tobacco roots and in cultured tobacco cells (Goossens et al, 2003; Katoh et al., 2005). Promoter regions of ∼250 base pairs from three PMT genes from Nicotiana sylvestris could confer jasmonate-responsive expression on a GUS reporter gene in transgenic hairy roots, showing that the jasmonate signal converges on relatively small promoter regions to confer transcriptional responses (Shoji et al., 2000). Two tobacco members of the AP2/ERF-domain transcription factor family called NtORC1 and NtJAP1 were shown to upregulate the activity of the tobacco PMT promoter in transient assays in tobacco protoplasts (De Sutter et al., 2005). Together the transcription factors caused a strong synergistic activation of the PMT promoter. NtORC1 is a close homologue of the Catharanthus AP2/ERF-domain transcription factor ORCA3. Both NtORC1 and NtJAP1 gene expression is induced by MeJA (Goossens et al., 2003).

Fig. 5
figure 5_8

Biosynthetic pathway of nicotine. Unbroken arrows indicate single enzymatic conversions and broken arrows indicate multiple enzymatic conversions. It is not known whether the N-methyl-pyrrolinium cation is coupled to nicotinic acid or a derivative of the latter. PMT; putrescine N-methyltransferase

Genetic studies using low-nicotine tobacco varieties demonstrated that the low-nicotine phenotype is caused by synergistic effects of two non-linked loci, called nic1 and nic2 (Katoh et al., 2005). The nic1nic2 double mutant has highly reduced nicotine contents (about 5% of wild type) and strongly decreased expression levels of nicotine biosynthesis genes. The genes corresponding to the nic loci have not been cloned yet.

In Nicotiana attenuata JA accumulates in response to attack by the herbivore Manduca sexta (tobacco hornworm) or in response to wounding and application of caterpillar oral secretions (a mimic of herbivore attack). This is likely caused by induction of jasmonate biosynthesis genes, since the NaLOX3 gene was induced by these treatments (Halitschke and Baldwin, 2003). Silencing the expression of NaLOX3 reduced nicotine accumulation in response to JA. Simultaneous silencing of NaJAR4 and NaJAR6 (the N. attenuata orthologues of AtJAR1) also reduced JA-responsive nicotine accumulation (Wang et al., 2008). Application of JA-Ile restored nicotine accumulation, indicating that JA-Ile is an important signalling molecule for nicotine production in Nicotiana attenuata.

Silencing of the COI1 gene in tobacco plants abolished the MeJA-responsive expression of nicotine biosynthesis genes including PMT, as well as MeJA-responsive nicotine accumulation (Shoji et al., 2008). The same report describes the isolation of three members of the tobacco JAZ gene family and their involvement in nicotine biosynthesis. The NtJAZ1–3 genes were induced by MeJA. For NtJAZ1 it was shown that the protein was rapidly degraded via the 26S proteasome in response to MeJA, whereas a derivative lacking the C-terminal Jas domain was stable. Overexpression of C-terminal deletion derivatives of NtJAZ1 or NtJAZ3 abolished MeJA-responsive PMT gene expression as well as nicotine accumulation.

These observations show that MeJA-responsive nicotine biosynthesis is controlled by the jasmonate receptor COI1 and depends on degradation of members of the JAZ repressor family (Fig. 6). There are no published data yet about the nature of the transcription factor(s) repressed by the JAZ proteins in tobacco, but in Fig. 6 we speculate that it is the tobacco homologue of AtMYC2 and CrMYC2. We also hypothesize that this NtMYC2 transcription factor controls the MeJA-responsive expression of the NtORC1 and NtJAP1 genes, which in turn are hypothesized to control the MeJA-responsive expression of the nicotine biosynthesis genes.

Fig. 6
figure 6_8

Model for jasmonate signal transduction leading to the expression of tobacco alkaloid biosynthesis genes. As depicted in Fig. 1, JA-Ile enhances the interaction between NtCOI1 and NtJAZ, leading to degradation of the latter proteins. NtMYC2 then activates transcription of the genes encoding the AP2/ERF-domain transcription factors NtORC1 and NtJAP1, which in turn activate the expression of tobacco alkaloid biosynthesis genes. NtMYC2 also activates transcription of NtJAZ genes as part of a negative feedback loop. The elusive NIC1 and NIC2 genes may encode NtMYC2 and/or NtORC1 and NtJAP1 as indicated by the question mark. The position of NtMYC2 in this signal transduction pathway is hypothetical as indicated by the question mark. Solid lines indicate interactions between proteins and broken lines indicate interactions between proteins and genes

7 Conclusion

We have given an overview of the current state of understanding of jasmonate signalling regulating alkaloid biosynthesis in tobacco and in Catharanthus roseus. For both species, certain elements in the models in Figs. 4 and 6 have not yet been experimentally confirmed. For the Catharanthus model, the involvement of the jasmonate receptor COI1 has not been experimentally confirmed. Given the conservation of COI1 as a jasmonate receptor in Arabidopsis, tomato, tobacco and Nicotiana attenuata, the position of COI1 in the jasmonate signal transduction pathway in Catharanthus seems highly probable. In tobacco, especially the identities and roles of transcription factors need more solid experimental confirmation. It will be interesting to see whether other secondary pathways regulated by JAs in different plant species are regulated in a similar manner.