CRISPR/Cas9-mediated disruption of ALLENE OXIDE SYNTHASE results in defective 12-oxo-phytodienoic acid accumulation and reduced defense against spider mite (Tetranychus urticae) in liverwort (Marchantia polymorpha)

Abstract

Allene oxide synthase (AOS) is a key enzyme involved in the biosynthesis of 12-oxo-phytodienoic acid (OPDA) and jasmonic acid and plays an important role in plant defense against herbivore attacks. In the liverwort, Marchantia polymorpha, we previously identified cytosol-type MpAOS1 and chloroplast-type MpAOS2 that show AOS activities. However, there is no direct evidence to show the subcellular localization of MpAOSs and their contribution to plant defense via OPDA production in M. polymorpha. In this study, we generated M. polymorpha mutants, with the MpAOS1 and MpAOS2 genes disrupted via CRISPR/Cas9-mediated genome editing; the loss of OPDA production was analyzed in double-knockout mutants. On AOS mutants, the survival rate and oviposition of spider mites (Tetranychus urticae) increased relative to those on wild-type plants. Overall, these findings suggest that defense systems via OPDA-signaling pathways in response to spider mites have been established in M. polymorpha.

The liverwort (Marchantia polymorpha) belongs to Bryophyta, which is a sister to the division containing vascular plants, namely, Tracheophyta (Morris et al. 2018). Owing to the environmental differences between the hydrosphere and land, early terrestrial plants had to adapt to various abiotic and biotic stresses. One adaptation mechanism that plants have developed is their ability to produce phytohormones, including 12-oxo-phytodienoic acid (OPDA), dinor-OPDA, and jasmonic acid (JA). OPDA, which is biosynthesized in the chloroplast through the octadecanoid pathway, and its converted product JA are known to be involved in defense in vascular plants (Wasternack and Strnad 2018). Unlike vascular plants, M. polymorpha produces OPDA/dinor-OPDA but not JA as a stress-signaling molecule (Monte et al. 2019, 2018). In addition, it has been reported that in M. polymorpha, salicylic acid and oxylipins, including OPDA, are antagonistic during plant–microbe interactions (Matsui et al. 2020). Furthermore, it is also known that terpenoid-rich oil bodies in M. polymorpha are involved in plant defense against arthropod herbivores such as pill bugs (Kanazawa et al. 2020; Romani et al. 2020). In our previous study, we found that M. polymorpha has two allene oxide synthases, MpAOS1 (Mp3g21350) and MpAOS2 (Mp5g16260), which convert 13-hydroperoxide of linolenic acid to OPDA in a coupled reaction with allene oxide cyclase (Koeduka et al. 2015). MpAOS2 has an extended sequence on its N-terminal (66 amino acid residues), responsible for chloroplast localization, whereas such a sequence is absent in MpAOS1 (Koeduka et al. 2015). In this study, we determined the subcellular localization of MpAOS1 and MpAOS2 in M. polymorpha cells and generated OPDA-deficient mutants of M. polymorpha through the CRISPR/Cas9-mediated disruption of MpAOS genes. In addition, we also investigated OPDA accumulation and defense traits toward spider mites (Tetranychus urticae).

First, to determine the subcellular localization of MpAOS1 and MpAOS2, full-length MpAOS1 and MpAOS2 cDNAs were translationally fused with a Citrine gene using Gateway cloning technology (Invitrogen, CA, USA) and constitutively expressed under the control of Cauliflower mosaic virus (CaMV) 35S promoter in M. polymorpha cells (Supplementary Figure S1A). The MpAOS-Citrine transgenic M. polymorpha plants were generated by Agrobacterium-mediated transformation methods (Kubota et al. 2013). The cells of the MpAOS1-Citrine transgenic line showed Citrine fluorescence in the cytosol, whereas the Citrine signals were detected clearly in the chloroplasts in the MpAOS2-Citrine transgenic line (Supplementary Figure S1B). This result indicates that MpAOS1 and MpAOS2 are localized to the cytosol and chloroplasts of M. polymorpha, respectively. Next, to explore the influence of AOS expression on OPDA production and the defense response of M. polymorpha to herbivore attacks, we examined the effects of the complete loss of MpAOS activity in M. polymorpha by targeting both cytosol-localized MpAOS1 and chloroplast-localized MpAOS2 via CRISPR/Cas9-mediated genome editing. The gRNA was designed near the initiation codons of MpAOS1 and MpAOS2 genes by using our in-house database (see Supplementary Table S1 for primers) and cloned into the pMpGE_En03 binary vector carrying Cas9 to generate the CRISPR/Cas9 construct in pMpGE010 (Sugano et al. 2018). The CRISPR/Cas9 vectors were transformed into Agrobacterium tumefaciens strain GV3101. For a single-knockout mutant, M. polymorpha sporelings were transformed through incubation with Agrobacterium carrying the CRISPR/Cas9 vectors for either MpAOS1 or MpAOS2. To generate the double-knockout mutants, the two Agrobacterium solutions were mixed before transformation. The transformants were obtained by hygromycin selection, as previously described (Ishizaki et al. 2016; Kubota et al. 2013). PCR-amplified fragments of MpAOS1 and MpAOS2, including the target site, were sequenced to identify the mutations by the method described previously (Supplementary Table S1; Sugano and Nishihama 2018). Sequence analysis indicated that the edited plants had indels and/or nucleotide substitutions at the target site (Figure 1A). Although the accumulation levels of MpAOS1 and MpAOS2 transcripts in the mutants was not fully abolished (Supplementary Figure S2), gene editing of the protein-coding region resulted in frameshifts, presumably leading to truncated and non-functional MpAOS proteins (Supplementary Table S2). It is likely that both MpAOS1 and MpAOS2 were fully inactivated in the Mpaos1 Mpaos2 mutants.

Figure 1. CRISPR/Cas9-mediated disruption of MpAOSs in Marchantia polymorpha. (A) Schematic diagram of the target sites in MpAOS1 and MpAOS2. Sequencing results from wild-type (WT, Tak-1) and three independent Mpaos1 Mpaos2 mutants of M. polymorpha are presented. Mutations based on insertion (or substitution) and deletion are indicated in blue and red, respectively. Numbers indicate the number of modified nucleotides. (B) Accumulation of 12-oxo-phytodienoic acid (OPDA) in wild-type and Mpaos1 Mpaos2 mutants of M. polymorpha. The values are the mean±SD (n=3). (C) LC-MS/MS chromatograms of OPDA extracted from wild-type and Mpaos1 Mpaos2 mutants of M. polymorpha. The MS was used in the negative ion mode, and ions were detected by multiple reaction monitoring (m/z; 290.9/165.2).

We then extracted OPDA and analyzed its level by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The thalli of M. polymorpha (approximately 500 mg), which were grown on half-strength Gamborg’s B5 agar medium for 2 weeks and frozen in liquid nitrogen, were extracted with 5 ml of ethyl acetate. The concentration of OPDA was analyzed by LC-MS/MS according to the method described by Mwenda et al. (2015). Three independent experiments were conducted using biological replicates. D₂-JA (25 ng per sample) was used as an internal standard for OPDA quantification. The LC-MS/MS analysis of OPDA in the three mutants revealed that the Mpaos1 Mpaos2 mutants did not show OPDA accumulation (Figure 1B, C). In contrast, single-knockout mutants of either MpAOS1 or MpAOS2 did not abolish OPDA accumulation in the transgenic liverwort. Unexpectedly, the OPDA level was significantly increased in Mpaos1 single-knockout mutants (Supplementary Figure S3). It seems that OPDA production by MpAOS2 was boosted by an unknown mechanism in the absence of MpAOS1. These data suggest that in the thallus, both pathways, localized in the cytosol and chloroplast, contribute to OPDA accumulation, although it is well known that a chloroplast-localized AOS is a key enzyme for producing OPDA and JA.

While previous studies on OPDA-deficient Physcomitrella patens mutants with disrupted PpAOS1 and PpAOS2 genes showed no distinct developmental defects (Luo et al. 2020), we observed a slight delay in thallus growth in the Mpaos double-knockout mutants, compared with that in the wild-type, after the gemmae of the double-knockout mutants were grown on half-strength Gamborg’s B5 agar medium for 3 weeks (Supplementary Figure S4). As the phenotypes of the mutants were almost the same, the Mpaos1-3 Mpaos2-3 line was utilized to investigate the influence of MpAOS disruption in response to a mite feeding assay.

As OPDA is known to play a central role in plant defense against herbivore attacks, we investigated whether MpAOS disruption affects the oviposition and survival rate of spider mites (T. urticae). For bioassays on the effects of MpAOS genes on the oviposition and survival rates of female spider mites, one of the females was released on a thallus (approximately 15 mm×15 mm) of the Mpaos1 Mpaos2 mutants or the wild-type, each growing from a gemma for 15 days, on cotton wool soaked in water in a plastic petri dish (diameter 35 mm, height 10 mm). After 7 days, we checked the survival of the females and counted their eggs on the thallus. The thallus of the Mpaos1 Mpaos2 mutants was infested with female T. urticae. There was a notable effect of MpAOS disruption on the survival rate of T. urticae, which was more pronounced on the thallus of the Mpaos1 Mpaos2 mutants relative to that on the wild-type, Tak-1 (Figure 2A). The oviposition of T. urticae was also affected by MpAOS disruption. The number of eggs laid by T. urticae was significantly higher on the thallus of Mpaos1 Mpaos2 mutants than on the thallus of the wild-type (Figure 2B). Although dinor-OPDA accumulation was not measured in this study due to the difficulty of our analytical techniques, it is likely that MpAOS disruption in the mutants also affects the levels of dinor-OPDA, as with OPDA production. These results suggest that T. urticae is affected by OPDA and/or dinor-OPDA-mediated defense systems that are controlled by MpAOSs. MpAOS-overexpressing lines should be further analyzed to reveal the mechanisms of the OPDA-mediated defense.

Figure 2. Effect of CRISPR/Cas9-mediated disruption of MpAOSs in Marchantia polymorpha on spider mite Tetranychus urticae survival rate and egg production. (A) Survival rate of T. urticae females (Fisher’s exact test). (B) Number of eggs laid by T. urticae females (Tukey–Kramer method after Box-Cox transformation). As the dataset for the number of eggs had 0 values, 0.5 was added to all values in the dataset before Box-Cox transformation. Different letters above the bars indicate significant differences (p<0.05). The number of replications for the wild-type and the Mpaos1 Mpaos2 mutants was 30 and 29, respectively.

Oil body-mediated protection of M. polymorpha against pill bugs has been demonstrated (Kanazawa et al. 2020; Romani et al. 2020), whereas the OPDA-mediated defense systems against attack by T. urticae have not been reported. In the OPDA-insensitive Mpmyc mutants, the herbivore Spodoptera littoralis did not induce the accumulation of sesquiterpenes, including thujopsene, β-chamigrene, and cuparene, unlike the wild-type (Peñuelas et al. 2019). Accordingly, it might be possible that chemical defense by sesquiterpene induction is partially accountable for the bioassay results of the wild-type and Mpaos1 Mpaos2 mutants. Moreover, it seems that OPDA produced by MpAOSs is responsible for the induced biosynthesis of secondary metabolites, including sesquiterpenes. Therefore, it is likely that non-vascular plants have acquired defense systems via OPDA-signaling pathways in response to mite feeding. The present findings will improve our understanding of adaptive signaling in response to biotic stress in terrestrial plants, as well as the physiological roles of MpAOS expression in herbivore-attack adaptation in M. polymorpha.