The fungal phytotoxin lasiojasmonate A activates the plant jasmonic acid pathway

The fungal jasmonate lasiojasmonate A functions as a source of jasmonic acid that activates the JA-isoleucine pathway in Arabidopsis.

Synthesis of phytohormones and phytohormone mimics is a common strategy of several pathogens to unbalance the plant immune system . The most-studied example is probably coronatine (COR), which is produced by some strains of Pseudomonas syringae. COR mimics the bioactive JA-Ile hormone by directly binding to the JA-Ile receptor COI1-JAZ (Brooks et al., 2005;Sheard et al., 2010). Lasiodiplodia mediterranea and other fungi also accumulate oxylipins such as OPDA, JA, and JA conjugates (Miersch et al., 1993;Andolfi et al., 2014;Chanclud and Morel, 2016;Fonseca et al., 2018). For example, Botryodiplodia theobromae, the causal agent of dieback disease of several plants, produces JA as a phytotoxin (Miersch et al., 1987;Gimenez-Ibanez et al., 2016). Fusarium oxysporum not only produces JA, but some strains also accumulate jasmonate conjugates such as JA-Ile and JA-Leu (Miersch et al., 1999;Brodhun et al., 2013;Cole et al., 2014). To promote infection, the activity of these jasmonates from F. oxysporum requires COI1, the plant JA-Ile receptor; this suggests a role for fungal JA-conjugates as virulent effectors (Cole et al., 2014). The rice blast fungus Magnaporthe oryzae was recently shown to convert natural JA into hydroxylated JA (12OH-JA), an inactive product of JA catabolism (Patkar et al., 2015). Although the activity of fungus-produced 12OH-JA is still not clear, secretion of this hydroxylated compound during infection to weaken plant defences suggests a role as a virulent factor (Patkar et al., 2015).
Although several JA derivatives have been isolated as fungal metabolites, their mode of action has not yet been determined. Here, we studied the effect of the natural fungal JA derivative LasA on plant JA-regulated responses, to define its mode of action in planta. LasA triggers many JA-regulated plant responses, including JAZ degradation, activation of JA gene expression, and growth inhibition. These effects in planta are dependent on JAR1 and COI1, which shows that plant conversion of LasA into JA, generation of JA-Ile, and its recognition by the JA-Ile receptor complex are necessary for the fungal LasA activity.

Plant material and growth conditions
Arabidopsis thaliana Col-0 was the genetic background of wild-type and mutant lines used in this study. Knock-out lines for coi1-1 (Xie et al., 1998), coi1-30 (Yang et al., 2012), coi1-2 (Xu et al., 2002), jar1-1 (Staswick and Tiryaki, 2004), myc2/jin1-2 (Lorenzo et al., 2004), and jai3-1 (Chini et al., 2007) have been described previously. Seeds were surface-sterilized by the chlorine gas method (in the presence of bleach and HCl for 3 h) and stratified for 2-3 d at 4 °C in the dark. All of the seedlings were grown under a 16-h light/8-h dark cycle at 21 °C (photosynthetic photon flux density approximately 100 µmol m -2 s -1 ). For root-growth inhibition assays, 10 to 30 seeds of each line were germinated for 10 d, with or without the presence of 50 µM jasmonic acid, 50 µM LasA, or 0.5 µM COR (Reveglia et al., 2018). Images were taken with a Nikon D1-x digital camera and root length was estimated using the ImageJ software (https:// imagej.nih.gov/ij/). Data were analysed by one-way ANOVA/Tukey HSD post hoc tests. Four independent biological replicates (10-30 seedlings each) were measured for each sample with similar results. Experiments were repeated four times with similar results.

Anthocyanin measurement
Between 10 to 30 seeds of each line were germinated for 10 d, with or without the presence of 50 µM jasmonic acid, 50 µM LasA, or 0.5 µM COR. Between four to ten seedlings were weighed and pooled for each biological replicate; three independent pools were measured for each line. Seedlings were soaked overnight in 1 ml of hydrochloric acid in aqueous methanol (HCl 0.5 N; methanol 80%, v/v). The reference blank was obtained by adding half of the sample solution to a peroxide reagent (one part 30% hydrogen peroxide in nine parts of methanolic hydrochloric acid, 5:1 v/v, 3N), whereas the other half of the solution was diluted with one volume of methanolic hydrochloric acid (5:1 v/v, 3N). Absorbance was measured at 530 nm, and anthocyanin content was expressed as (A530 sample -A530 blank) per g fresh weight. Data were analysed by one-way ANOVA/Tukey HSD post hoc tests. Experiments were repeated twice with similar results.
Yeast two-hybrid system All constructs were previously reported by Chini et al. (2007). pGAD (prey) and pGBK (bait) plasmids were co-transformed in Saccharomyces cerevisiae AH109 cells using standard heat-shock protocols (Chini et al., 2007). Successfully transformed colonies were selected 3 d after transformation on yeast synthetic drop-out medium lacking Leu and Trp (-2 medium). Yeast colonies were then grown in selective -2 liquid medium for ~5 h and cell density adjusted to 3 × 10 7 cells ml -1 (OD 600 =1). Cell suspensions (5 µl) were plated on yeast synthetic drop-out medium lacking Ade, His, Leu, and Trp (-4 medium) to evaluate protein interactions. Where indicated, the medium was supplemented with 20 µM coronatine and/or 100 µM LasA. Plates were incubated (2-6 d; 28 °C) and images were acquired with a Nikon D1-x digital camera.

pJAZ2-GUS expression assays
The pJAZ2:GUS marker line was described previously by Gimenez-Ibanez et al. (2017). For pJAZ2:GUS expression experiments, seeds were germinated as described above and 6-d-old seedlings were treated with either liquid Johnson medium (control), 5 µM JA, or 100 µM LasA (2 h); GUS was stained as described above and incubated overnight at 37 ºC. Images were acquired with a Nikon D1-x camera. Analyses were performed using 10-20 plants per sample and the experiment was repeated three times with similar results.
Quantitative RT-PCR Quantitative RT-PCR was performed using biological samples of tissue pooled from 10-20 wild-type seedlings. RNA was extracted and purified using Trizol reagent (Invitrogen) followed by use of a High Pure RNA isolation kit (Roche), including DNase digestion to remove genomic DNA contamination. cDNA was synthesized from 1 μg total RNA with a high-capacity cDNA reverse transcription kit (Applied Biosystems). For gene amplification, 4 μl from a 1:10 cDNA dilution was added to 4 μl of EvaGreen® qPCR Mix Plus (Solis BioDyne) and gene-specific primers as previously described . Quantitative PCR was performed in 96-well optical plates in a HT 7900 Real Time PCR system (Applied Biosystems) using standard thermocycler conditions (an initial hold at 95 °C for 10 min, followed by a two-step SYBRPCR program of 95 °C for 15 s and 60 °C for 60 s for 40 cycles). Relative expression values given as the means of three or four technical replicates relative to the basal wild-type control using ACT8 as the housekeeping gene. Data were analysed using unpaired Student's t-tests.

Effect of LasA in planta
(-)-JA, its methyl ester, and lasiojasmonate A (Fig. 1A, 1-3) were purified from the organic extracts of Lasidiplodia mediterranea (BL 101 strain) culture filtrates and identified by comparing their physical and spectroscopic properties with those previously reported . To assess its effects in planta, wild-type (WT) Arabidopsis plants were grown with or without fungal LasA. Similar to JA, LasA induced growth inhibition and anthocyanin accumulation ( Fig. 1B-F; Supplementary Fig. S1 at JXB online). To test whether the activity of this JA derivative required in planta conjugation of JA to Ile, JA-Ile perception, and the JA signalling pathway, we also grew seeds of jar1-1 (impaired in JA-Ile conjugation), coi1-2 (weak allele of the JA-Ile receptor), and jai3-1 and jin1-2 (both JA signalling mutants) in the presence of LasA. LasA showed an effect similar to that of JA in all genotypes, strongly inhibiting growth of WT plants but only partially inhibiting jar1-1, jai3-1, jin1-2, and coi1-2 ( Fig. 1B-F; Supplementary Table S1). LasA induced accumulation of the defence secondary metabolites anthocyanins, and this response required a functional JA pathway ( Supplementary Fig. S1).
These results show that fungal LasA triggers JA-regulated responses, and that this LasA-induced effect requires the canonical JA pathway.

Effect of LasA on JA-mediated plant transcriptional activation
As LasA triggers physiological responses mediated by JA, we evaluated whether it could activate JA-dependent gene expression. We analysed the transcriptional changes of wellknown JA-marker genes induced by LasA, including several JAZ genes as well as JA-biosynthetic genes such as LOX3, AOS, AOC1, and OPR3. Exogenous treatment of LasA triggered the expression of all tested JA-marker genes within 1 h ( Fig. 2A; Supplementary Fig. S2).
To provide further evidence of the role of LasA in the activation of JA signalling, we assessed the effect of LasA on the JA-inducible marker line pJAZ2:GUS. As expected, LasA activated expression of the marker line, although at a lower level than JA (Fig. 2B). To assess whether this activation required JA-Ile perception by COI1, we tested activation of the marker line in the coi1-30 background. Similar to JA, fungal JA derivatives did not activate pJAZ2:GUS expression in coi1-30 mutant plants (Fig. 2B).
These data indicate that LasA induces expression of JA-regulated genes in a COI1-dependent manner.

Effect of LasA on the JA-Ile co-receptor COI1-JAZ complex
LasA triggers physiological responses mediated by JA in a COI1-and JAR1-dependent manner (Fig. 1), suggesting that it is first converted to JA, then conjugated to form JA-Ile that would in turn trigger JAZ protein degradation. To support this hypothesis, we evaluated the LasA effect on hormonemediated formation of the COI1-JAZ co-receptor complex in a yeast-two-hybrid assay (Y2H; Chini et al., 2009;Chini, 2014). LasA did not promote COI1-JAZ3 interaction in the absence of the hormone (Fig. 3A). To assess a possible LasA effect on yeast growth and Y2H detection of protein-protein interactions, we also assessed LasA activity on formation of the complex between JAZ3 and the transcription factor MYC2. LasA did not alter the JAZ3-MYC2 interaction. In the presence of COR, COI1 binds to JAZ3; LasA nonetheless did not affect the COR-induced COI1-JAZ3 interaction (Fig. 3B). These findings show that LasA does not directly interfere with the JA-Ile receptor complex COI1-JAZ.

Effect of LasA on JAZ turnover
LasA induced JA-dependent gene expression without interfering directly with the COI1-JAZ receptor; therefore, LasA might be converted to JA and then conjugated to form JA-Ile, which in turn would trigger JAZ protein degradation. To test this hypothesis, we evaluated the effect of LasA on JAZ repressor degradation by studying JAZ1-GUS turnover in the presence of LasA. Similar to JA or COR treatments, LasA induced JAZ1 degradation (Fig. 4A). To determine whether JA must be conjugated into JA-Ile for LasA activity, we assessed the effects of LasA on JAZ1 stability in jar1-1 mutants, which have partially impaired JA-Ile conjugation. LasA induced a partial JAZ1-GUS degradation, similar to Yeast cells co-transformed with pGAD-JAZ3 (prey) and pGBK-COI1 (bait) were selected and subsequently grown on synthetic drop-out medium lacking Leu and Trp (-LW) as the transformation control, or on selective media lacking Ade, His, Leu, and Trp (-AHLW) to test protein interactions. COI1 interaction with JAZ3 was detected only in the presence 20 µM coronatine (B). LasA (at 100 µM) did not alter COI1/JAZ interactions in the presence of COR. As a control, the interaction between JAZ3 (bait) and MYC2 (prey) was also assessed (A, lower panels). LasA did not promote COI1/JAZ3 interaction, nor alter the MYC2/JAZ3 interaction (B, lower panels). that promoted by JA (Fig. 4A, B). In addition, we studied the effect of LasA in coi1-1 plants that have loss of JA-Ile receptor function, in order to verify the requirement for JA-Ile perception for LasA activity. LasA did not induce JAZ1-GUS degradation in the complete loss-of-function coi1-1 mutant (Fig. 4A, B) and, as expected, JA also did not promote JAZ1-GUS degradation in this mutant. The JA-Ile analogue COR acts as a ligand of the COI1-JAZ receptor and activates JAZ degradation and JA responses independently of JAR1 (Sheard et al., 2010) and indeed, our results showed that COR induced JAZ1 degradation in a COI1-dependent, JAR1-independent manner (Fig. 4A). Finally, LasA did not prevent hormone-triggered degradation of JAZ1 during concurrent JA and LasA treatment ( Supplementary Fig. S3).
These results show that LasA induces JAZ degradation, and that its activity requires JAR1-mediated JA conversion into JA-Ile, as well as COI1-dependent perception of JA-Ile.

Discussion
Jasmonates are signalling molecules that govern plant responses to stress. Jasmonic acid is the direct precursor of the bioactive plant hormone JA-Ile, the endogenous ligand of the plant receptor complex. Several fungi produce JA and JA derivates; phytotoxic activity is reported for most of these compounds, although a precise molecular mode of action has yet to be proposed. Here, we addressed the activity of the natural JA derivate LasA, which is produced by the pathogenic fungus Lasiodiplodia mediterranea . Our results showed that LasA activated the plant JA pathway and that its activity required JAR1 and COI1, key proteins required for JA conversion into JA-Ile and for JA-Ile perception, respectively. Because LasA is a fungal JA furanonyl ester, we propose that this naturally occurring fungal compound can be catabolized to release JA, which is in turn converted into JA-Ile by JAR1. In support of this hypothesis, LasA activity required JAR1 for JA conjugation to Ile, as well as recognition by the JA-Ile perception complex. LasA might therefore function as an inactive JA conjugate that can be hydrolysed to liberate the direct JA-Ile precursor JA.
A cleaving amidohydrolase activity on JA conjugates has been described for the JA-producing fungus Botryodiplodia theobromae (Hertel et al., 1997). This enzymatic activity might be conserved among fungi, which would explain the detection of notable amounts of JA in several species with only minimal accumulation of JA-conjugated molecules (Chanclud and Morel, 2016;Lo Presti et al., 2015;Eng et al., 2016;Fonseca et al., 2018); however, the fungal cleaving amidohydrolase activity on JA conjugates might mask identification of additional natural JA-conjugates under in vitro conditions. Plants have also evolved amidohydrolase activity; the plant enzyme IAR3/ILL6 deconjugates JA-Ile into JA (Woldemariam et al., 2012;Bhosale et al., 2013;Widemann et al., 2013). This process might accept different substrates, since IAR3/ILL6 can also deconjugate 12OH-JA-Ile into 12OH-JA (Widemann et al., 2013). In addition, several members of this amidohydrolase family release free auxin (IAA) by cleaving IAA-amino acid conjugates, with each hydrolase deconjugating a different subset of the conjugates (LeClere et al., 2002). Finally, the expression of several amidohydrolase genes is induced by auxins and jasmonate, strengthening the case for the involvement of amidohydrolase in hormone signalling (LeClere et al., 2002;Zhang et al., 2016). Cleaving amidohydrolase activity of molecule-amino acid conjugates is thus a fairly common mechanism in nature, and both fungi and plants might be able to hydrolyse JA-conjugates such as LasA.
LasA activation of JA responses is dependent on the JA-Ile co-receptor COI1. A similar requirement for the plant COI1 JA-Ile receptor was reported for the virulent activity of fungal JA-conjugates produced by Fusarium oxysporum (Miersch et al., 1999;Brodhun et al., 2013;Cole et al., 2014). These data suggest that different fungi have evolved distinct JA-conjugates with similar biological activity dependent on the plant COI1 JA-Ile receptor.
Jasmonic acid regulates several plant responses, including plant defences against necrotroph pathogens, fungi, and herbivores (Wasternack and Hause, 2013;Goossens et al., 2016;Guo et al., 2018). On the other hand, activation of the JA pathway weakens the defences against biotrophic and hemibiotrophic pathogens triggered by salicylic acid (SA) (Robert-Seilaniantz et al., 2011). Hence, some strains of Pseudomonas syringae produce a mimic of the bioactive JA-Ile hormone, coronatine, to activate the JA pathway, which in turn inhibits the SA-dependent defences required for P. syringae resistance (Cui et al., 2005;Robert-Seilaniantz et al., 2011). A biotrophic or hemibiotrophic infection stage for L. mediterranea has not been reported to date, providing a case against the hypothesis that L. mediterranea produces LasA to activate JA-dependent responses to inhibit SA defences.
The suggested mode of action for fungal LasA-that it functions as a pool of inactive conjugated JA that is converted to active JA, and hence activates plant JA responses including defence against necrotrophic fungi-might seem counterintuitive at first glance. However, necrotrophic fungi induce plant cell death, which is beneficial for fungal growth and proliferation (Chanclud and Morel, 2016;Lo Presti et al., 2015). We therefore propose that fungal production of LasA, and possibly of additional JA derivates, would be spatio-temporally regulated to activate JA-mediated senescence and cell death only in specific conditions. For example, to avoid inducing plant defences, fungi would not produce LasA in early infection stages, whereas production and subsequent induction of plant cell death would occur in late stages in order to facilitate fungal propagation and infection. Future analysis of Lasiodiplodia mutants with impaired LasA production will test this hypothesis.

Supplementary data
Supplementary data are available at JXB online. Fig. S1. Accumulation of anthocyanins in response to exogenous treatment with JA, LasA, or COR. Fig. S2. Effects of LasA on JA-regulated transcriptional activation. Fig. S3. Effect of LasA on JA-mediated JAZ degradation. Table S1. Data for root growth measurements reported in Fig. 1F.