The lre‐miR159a‐LrGAMYB pathway mediates resistance to grey mould infection in Lilium regale

Abstract Grey mould is one of the most determinative factors of lily growth and plays a major role in limiting lily productivity. MicroRNA159 (miR159) is a highly conserved microRNA in plants, and participates in the regulation of plant development and stress responses. Our previous studies revealed that lre‐miR159a participates in the response of Lilium regale to Botrytis elliptica according to deep sequencing analyses; however, the response mechanism remains unknown. Here, lre‐miR159a and its target LrGAMYB gene were isolated from L. regale. Transgenic Arabidopsis overexpressing lre‐MIR159a exhibited larger leaves and smaller necrotic spots on inoculation with Botrytis than those of wild‐type and overexpressing LrGAMYB plants. The lre‐MIR159a overexpression also led to repressed expression of two targets of miR159, AtMYB33 and AtMYB65, and enhanced accumulation of hormone‐related genes, including AtPR1, AtPR2, AtNPR1, AtPDF1.2, and AtLOX for both the jasmonic acid and salicylic acid pathways. Moreover, lower levels of H2O2 and O2- were observed in lre‐MIR159a transgenic Arabidopsis, which reduced the damage from reactive oxygen species accumulation. Taken together, these results indicate that lre‐miR159a positively regulates resistance to grey mould by repressing the expression of its target LrGAMYB gene and activating a defence response.

to function in many plant biological and metabolic processes. Many plant miRNAs have been reported to respond to abiotic challenge, suggesting a broader involvement of miRNAs in defence (Jones-Rhoades and Bartel, 2004;Phillips et al., 2007;Shanfa et al., 2010).
For example, miR398 expression was down-regulated by oxidative stresses, and the expression of its target genes Cu/Zn SUPEROXIDE DISMUTASES was increased to protect plants from oxidative damage during SO 2 exposure . In addition, the overexpression of miR393a causes plants to become tolerant to drought stress, salt stress, and heat stress . Moreover, miRNAs have also been shown to be pivotal molecules in plant-pathogen interactions. For instance, miR396 was confirmed to target a set of transcription factors in the GROWTH-REGULATING FACTOR (GRF) family in response to the necrotrophic fungi Plectosphaerella cucumerina and Botrytis cinerea, and the hemibiotrophic fungi Fusarium oxysporum f. sp. conglutinans and Colletotrichum higginsianum. The MIM396 plant, in which miR396 activity is reduced, was more resistant to these fungal infections, whereas the overexpression of MIR396 increased plant susceptibility to fungal infections (Soto-Suárez et al., 2017). Moreover, slmiR482e-3p, which targets FRG3 (a novel R gene) at the transcript level, is necessary for resistance to tomato wilt disease (Ji et al., 2018). Several miRNAs have been reported to respond to B. cinerea. In tomato, miR160 and miR171a were up-regulated, and miR169 was down-regulated after inoculation with B. cinerea (Jin et al., 2012). In addition, miR394 was verified as a negative regulator of Arabidopsis resistance to B. cinerea by targeting LEAF CURLING RESPONSIVENESS (LCR) (Tian et al., 2018).
Furthermore, large numbers of miRNAs were identified in response to grey mould in Solanum lycopersicum, Paeonia lactiflora, and Lilium regale through high-throughput sequencing (Jin and Wu, 2015;Zhao et al., 2015;Gao et al., 2017), which might be associated with resistance to Botrytis stress.
The miR159 family is a conserved miRNA that has been found in most land plants except bryophytes (Allen et al., 2007). miR159 has been found to negatively regulate GAMYB or GAMYB-like transcription factors, which activate gibberellin (GA)-responsive genes (Achard et al., 2004). In Arabidopsis, seven of the miR159 targets encode proteins similar to GAMYB, all of which share conserved putative miR159-binding sites of analogous complementarity, such as MYB33, MYB65, and MYB101 (Millar, 2005;Zheng et al., 2017).
The functional role of the miR159-MYB pathway has been analysed during seed germination floral development (Tsuji et al., 2006), and primary root growth (Xue et al., 2017). Despite this, more research has focused on the function of miR159 in response to multiple environmental stresses. In sugarcane, a progressive increase in miR159 transcripts was observed under short-term polyethylene glycol stress with a concomitant down-regulation of target MYB genes (Patade and Suprasanna, 2010). Similarly, root endophytic fungi induced the accumulation of miR159, enhancing the tolerance of drought stress in rice (Ehsan et al., 2017). In wheat, miR159 was up-regulated in leaves when challenged with Puccinia striiformis f. sp. tritici, resulting in a resistant phenotype through the regulation of TaMYB3 expression (Feng et al., 2013). The overexpression of miR159 caused increased resistance to powdery mildew in Arabidopsis (Sun, 2014). All of these indicate that miR159 is a potential positive regulator in plant response to a variety of biotic and abiotic stresses.
Previously, we systematically investigated the role of different miRNAs in the response of L. regale to B. elliptica and found that the miR159 family was significantly expressed after inoculation with B. elliptica (Gao et al., 2017). To further our knowledge about the functions of miR159, unravelling its role in response to Botrytis and the underlying molecular mechanisms, we carried out this study. The miR159a and target LrGAMYB genes were identified and characterized in L. regale, and transgenic Arabidopsis plants overexpressing lrepre-MIR159a were generated. Particularly interesting effects were observed in transgenic Arabidopsis plants overexpressing miR159a, which had markedly increased resistance to grey mould in comparison with wild-type controls. The target of lre-miR159a, LrGAMYB, is repressed in the transgenic plants. Our results reveal that the miR159a acts as a positive regulator of grey mould tolerance in Lilium.

| Identification of lre-miR159a in L. regale
We took advantage of L. regale, which shows resistance to grey mould, to clone lre-miR159a. The length of precursor lre-MIR159a was 219 nt, containing a 21 nt mature miR159a sequence at the 3′ end. Sequence alignment of the MIR159a precursor revealed 52.97% identity between Arabidopsis thaliana and L. regale, as shown in The mature miR159 sequence base conservation was analysed among 106 members, as shown in Figure 1d. The consensus mature miR159 sequence was 5′-UUUGGAUUGAAGGGAGCUCUA-3′ and shared high identity from the second to the 20th nucleotide. Mature lre-miR159a shared the same sequence as the consensus sequence, indicating that lre-miR159a might play a similar role as in other species.

| LrGAMYB is targeted by miR159a through cleavage
It is well known that plant miRNAs function through cleaving target genes. Therefore, identifying target genes of mRNAs is important to understand their specific contributions. In plants, miR159 was predicted to target MYBs with a strongly conserved miR159-binding site. To identify miR159 target genes in Lilium, psRNATarget was To confirm whether LrGAMYB genes are direct targets of lre-miR159a, a modified 5′ RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE) was conducted to examine the miR159-directed cleavage sites of LrGAMYB transcripts. Generated cleavage products were amplified and cloned into a vector. The 5′ end sequencing of the amplified products was sequenced in independent clones, suggesting that cleavage sites are located in the middle of the lre-miR159a complementary region. Cleavage occurred at an identical position, which corresponded to the 10th nucleotide position of the consensus mature miR159a sequence, in 8 of the 10 tested samples ( Figure 3c). Furthermore, L. regale plants were inoculated with B. elliptica to investigate the expression pattern of LrGAMYB using RT-qPCR analysis ( Figure 3d). This revealed that LrGAMYB was up-regulated in the initial B. elliptica infection and significantly reduced at later times. The expression level of LrGAMYB was negatively correlated with lre-miR159a in L. regale. These results confirm that LrGAMYB is an authentic target of lre-miR159a, and LrGAMYB is subject to miR159-mediated down-regulation.

| lre-miR159a positively regulates plant resistance to B. elliptica
To study the function of lre-miR159a and LrGAMYB in plant pathogens, we constructed two plasmids overexpressing lre-miR159a and LrGAMYB, respectively, and transformed these into the Arabidopsis

| The expression of hormone-related genes in lre-miR159a overexpression Arabidopsis after infection with B. cinerea
Phytohormone signalling pathways, such as salicylic acid (SA) and

| miR159a overexpression balances ROS homeostasis and increased resistance to Botrytis
Because one of the earliest defence responses in B. cinerea is reactive oxygen species (ROS) production (Asselbergh et al., 2007;van Kan, 2006)

| D ISCUSS I ON
The miR159 family is one of the ancient miRNA families in not only in monocotyledons and dicotyledons but also gymnosperms and ferns, such as Picea abies (Xia et al., 2015), Pinus taeda (Lu et al., 2007), and Selaginella moellendorffii (Axtell et al., 2007). Furthermore, the miR159 family differs in the number of mature and precursor genes among species. In Arabidopsis, this family is encoded by three genes and forms three mature miR159 members (miR159a, miR159b, and miR159c) located in different regions of the genome (Allen et al., 2007). Further research has indicated that a mir159a mir159b double mutant has pleiotropic morphological defects, including altered growth habits, curled leaves, small siliques, and small seeds. Neither mir159a nor mir159b single mutants displayed any of these traits, indicating functional redundancy (Allen et al., 2007;Alonso-Peral et al., 2010). In our previous research, five miR159 members were identified in L. regale through high-throughput sequencing; however, lre-miR159a was significantly more highly and differentially expressed than others members after inoculation with B. elliptica, suggesting that lre-miR159a might respond to B. elliptica in L. regale (Gao et al., 2017). Therefore, we chose lre-miR159a for further research in this study.
The results of sequence alignment analysis suggested that mature miR159a sequences are highly conserved in different species; however, the precursors have substantial differences. For instance, Z. mays and P. abies possess the most miR159 precursors with 11 genes, but 11 species hold only one. Furthermore, the phylogenetic tree showed that the evolutionary distance separating the branches of different species of miR159a precursors is long, and the sequence similarity is relatively low, suggesting varied sources of mature miR159a. pre-lre-MI159a formed an independent clade, suggesting it is different from other species. The highly conserved sequences of mature miR159 indicate that they may perform similar functions.
Studies have revealed that miR159 plays an important role in the response to stresses. For instance, miR159 responds to salinity stress (Kitazumi et al., 2015), drought stress (Mohsenifard et al., 2017), and heat stress (Li et al., 2016). Suppression of miR159 in plants resulted in short stature along with smaller stem, leaf, and grain size, which enhanced the hypersusceptibility to adverse environmental stress . Furthermore, altered accumulation of miR159a levels was observed during tomato leaf curl virus infection in tomato, indicating that miR159 behaves as an active factor in pathogen resistance (Koundal et al., 2010). In our previous study, we also found by high-throughput sequencing that lre-miR159a was involved in the response to Botrytis. miR159 mainly contributes to the regulation of plant development and stress by targeting MYB transcription factors (Millar, 2005;Xue et al., 2017). The GAMYB or GAMYB-like genes encode a highly conserved family of R2R3 MYB domain transcription factors implicated in GA signal transduction (Woodger et al., 2003). Studies have indicated that the GAMYB transcription factor superfamily is F I G U R E 6 The expression of hormone-related genes of wild-type (WT) and Arabidopsis overexpressing miR159a (OElre-miR159a) upon inoculation with Botrytis cinerea. The level of expression was normalized to the level of the AtActin gene. The normalized gene levels at 0 hr were arbitrarily set to 1. Each bar shows the mean ± SE of triplicate assays. * or ** indicates a statistically significant difference relative to the value at 0 hr for each gene at p < .05 or .01, respectively involved in plant development and metabolism (Alonso-Peral et al., 2010;Sheldon et al., 2001) and response to pathogens and abiotic stress (Yang et al., 2014;Butt et al., 2017). Moreover, studies have revealed that MYB transcriptome factors participate in response to B. cinerea. For instance, the overexpression of AtMYB44 in Arabidopsis results in a stronger ROS burst, greater cell death, and severe necrosis symptoms, which enhances susceptibility to B. cinerea (Shi et al., 2011). Furthermore, MYB46, thought to regulate secondary cell wall biosynthesis in the vascular tissue of the stem, functions as a disease susceptibility modulator to B. cinerea. The MYB46-mutant plants exhibited increased disease resistance to B. cinerea (Vicente et al., 2011). The results indicate that the molecular and physiological responses to Botrytis included ROS production and transcriptional responses.
A stronger response of JA-and SA-mediated defence genes was detected in the Botrytis-infected OEmiR159a plants (Figure 6), implying that overexpression of miR159a enhances transgenic plant defence by activating the hypersensitive response. An SA-dependent signalling pathway led to the expression of pathogenesis-related (PR) proteins, including AtNPR1, AtPR1, and AtPR2, thus contributing to resistance. AtPDF1.2, which is dependent on the JA pathway, was also induced and was highly expressed. In agreement with previous reports, the SA and JA signalling pathways interact extensively and cooperatively in response to Botrytis infection (Glazebrook, 2005).
B. cinerea is a nonspecific necrotrophic fungal pathogen that triggers plants to generate large amounts of ROS and induces local cell death to facilitate infection (Su et al., 2011;Pietrowska et al., 2015).
Previous studies have shown that the induction of H 2 O 2 in plant cells, accompanied by O − 2 generation, can promote programmed cell death in the host and the expansion of disease lesions to facilitate B. cinerea infection (Govrin and Levine, 2000;Patykowski, 2006;Asai and Yoshioka, 2009;Wan et al., 2015). During the early stages of Botrytis infection, the ROS burst can induce a defensive reaction. However, high H 2 O 2 levels can disturb redox homeostasis, trigger initiation of cell death, and facilitate necrotrophic pathogen attack of host plants. In our study, H 2 O 2 and O − 2 levels were lower in OEmiR159a Arabidopsis than in WT and OELrGAMYB Arabidopsis, most probably due to the overexpression of lre-miR159a. These results indicate that lre-miR159a plays a positive role in resistance to Botrytis by suppressing the target LrGAMYB gene.

We also found an interesting phenomenon in which OEmiR159a
Arabidopsis showed early flowering. Studies have indicated that miR159 is involved in floral organ development. For instance, miR159 expression regulates floral transition , flower development , and timing of flowering (Guo et al., 2017). Furthermore, miR159 is required for fruit set (da Silva et al., 2017), and the accumulation of miR159 could affect seedling development. These results suggest multiple regulatory networks of miRNAs that participate in plant growth and development. Further research is needed to verify the functions of these miRNAs and their regulatory networks.

| Sequence analysis of precursor miR159a from L. regale and prediction target genes
Total genomic DNA was extracted from leaves using the New Plant Genomic DNA Extraction Kit (Tiangen). The DNA sequence was amplified by PCR using primers designed based on the small-RNA library previously constructed by our laboratory (Gao et al., 2017), and the primers are shown in Table S1. PCR amplification was per-  (Dai and Zhao, 2011). The prediction of the lre-miR159a targets was based on the transcriptome of L. regale, which was constructed by our laboratory. A phylogenetic tree was constructed with MEGA 6 using the neighbour-joining method. The conserved domains of mature miR159a sequences were aligned using the Weblogo program with default parameters (http://weblo go.berke ley.edu/logo.cgi) (Crooks et al., 2004).

| Cloning the lre-miR159a-targeted GAMYB gene and sequence analysis
The lre-miR159a-targeted GAMYB partial sequence was deduced from the L. regale transcriptome. First, total RNA from L. regale leaves was extracted using EASYspin Plus RNA Kit (Aidlab). Full GAMYB cDNA fragments were then obtained with the SMARTer RACE 5′/3′ Kit (Takara) followed by the second round of nested PCR with specific outer and inner primers listed in Table S2. The full-length of GAMYB gene sequence was amplified with designed primers (Table S1) and cloned into the pLB vector. These GAMYB protein sequences of other plants were obtained from the NCBI database (https://www.ncbi. nlm.nih.gov/). Multiple alignments of these protein sequences were aligned using DNAMAN.

| RNA ligase-mediated rapid amplification of cDNA ends
To detect the miRNA-target cleavage site, RLM-RACE was conducted using the 5′-Full RACE Kit (Takara) with a 5′-RACE adaptor (Lu et al., 2017). Total RNA for RLM-RACE was obtained from L. regale leaves. The 5′-RACE outer primer and gene-specific outer primer (GSP1) were used for the first round of nested PCR, followed by the second round of nested PCR using the 5′-RACE inner primer and gene-specific inner primer (GSP2) ( Table S2).
Amplification products were gel purified, cloned into the pRACE vector, and sequenced.

| Plasmid construction and Arabidopsis transformation
To overexpress miR159a and LrGAMYB, the sequence of the precursor lre-miR159a and full cDNA fragments encoding LrGAMYB were amplified from the corresponding cloned vectors and then inserted downstream of the CaMV 35S promoter in pCAMBIA1301 (GenBank no. AF234297). The recombinant vectors were then introduced into Agrobacterium tumefaciens GV3101 and the Arabidopsis Col-0 WT was used for transformation by the floral-dip method.
Transformants were selected with 50 mg/L hygromycin B and confirmed by reverse transcription PCR. Three homozygous T 3 lines were established and used for the treatments. For the bioassay, discs of B. cinerea mycelia and conidia solution were both used for the inoculation of detached leaves.
Discs of mycelia were punched from the growing edge of colonies, and conidia were adjusted to a concentration of 5 × 10 5 conidia/ml for inoculation. Trypan blue staining for the presence of the fungal infection was performed as previously described .
After the decolorization treatment, the mycelium in tissue was observed with a microscope (BME, Leica).

| RNA insolation and RT-qPCR
Total RNAs were extracted and reverse transcribed using the First Strand cDNA Synthesis Kit (Toyobo). RT-qPCR was conducted in a total volume of 20 μl containing 10 μl SYBR Premix Ex Taq (Takara) using following programme: 5 min denaturation at 94 °C; followed by 30 cycles of 94 °C for 5 s, 60 °C for 20 s, and 72 °C for 20 s. The 18S rRNA and CLATHRIN genes were used as the reference genes for normalization in L. regale, and the AtActin gene was used for A. thaliana.
Each sample was performed in triplicate and the mean value of technical replicates was recorded for each biological replicate. The primers used are listed in Table S3.
For RT-qPCR analysis in Arabidopsis, inoculated leaves were harvested at 0, 24, and 48 hpi for RNA extraction.

| Statistical analysis
All data are presented as the mean ± SE and were subjected to analysis of variance according to Student's t test (*p < .05, **p < .01). Each data set was independently compared with the control data set to determine significant differences.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.

S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the Supporting Information section.
How to cite this article: Gao X, Zhang Q, Zhao Y-Q, Yang J, Jia