An RXLR effector PlAvh142 from Peronophythora litchii triggers plant cell death and contributes to virulence

Abstract Litchi downy blight, caused by the phytopathogenic oomycete Peronophythora litchii, results in tremendous economic loss in litchi production every year. To successfully colonize the host cell, Phytophthora species secret hundreds of RXLR effectors that interfere with plant immunity and facilitate the infection process. Previous work has already predicted 245 candidate RXLR effector‐encoding genes in P. litchii, 212 of which have been cloned and tested for plant cell death‐inducing activity in this study. We found three such RXLR effectors could trigger plant cell death through transient expression in Nicotiana benthamiana. Further experiments demonstrated that PlAvh142 could induce cell death and immune responses in several plants. We also found that PlAvh142 localized in both the cytoplasm and nucleus of plant cells. The cytoplasmic localization was critical for its cell death‐inducing activity. Moreover, deletion either of the two internal repeats in PlAvh142 abolished the cell death‐inducing activity. Virus‐induced gene silencing assays showed that cell death triggered by PlAvh142 was dependent on the plant transduction components RAR1 (require for Mla12 resistance), SGT1 (suppressor of the G2 allele of skp1) and HSP90 (heat shock protein 90). Finally, knockout of PlAvh142 resulted in significantly attenuated P. litchii virulence on litchi plants, whereas the PlAvh142‐overexpressed mutants were more aggressive. These data indicated that PlAvh142 could be recognized in plant cytoplasm and is an important virulence RXLR effector of P. litchii.

signalling by plant hormones and transcriptional regulation via transcription factors (Pitzschke et al., 2009;Wang et al., 2019b). The immune activation culminates in a series of physiological changes in the plant, such as reactive oxygen species (ROS) production, cell wall fortification, and the localized rapid cell death known as the hypersensitive response (HR) (Ingle et al., 2006;Franceschetti et al., 2017).
Oomycetes are a group of straminipilous organisms that are phylogenetically distant from true fungi (Göker et al., 2007;Beakes et al., 2012). Notably, phytopathogenic oomycetes are a constant threat to many important crops, rendering enormous crop yield losses globally (Tyler, 2007;Lamour et al., 2012;Fry et al., 2015). The phytopathogenic oomycetes could use the complement effectors, secreted into either apoplastic or cytoplasmic regions, as major virulence factors for successful infection and causing disease symptoms (Tyler et al., 2006;Wang and Wang, 2018a). Among them, the cytoplasmic RXLR effectors that contain an N-terminal signal peptide followed by the conserved Arg-any amino acid-Leu-Arg (RXLR) motif, are a large superfamily of virulence proteins in oomycetes (Rehmany et al., 2005;Whisson et al., 2007). The RXLR motif is located within 30 residues downstream of the secretion signal cleavage site and is frequently followed by a less conserved Asp-Glu-Glu-Arg (dEER) motif (Wawra et al., 2012). It is suggested that the RXLR-dEER motif is involved in translocating effector proteins from haustoria into host cells (Dou et al., 2008;Kale et al., 2010). RXLR effectors have become a focus for studying plant-pathogen interaction in the past decade, with numerous effector genes identified and characterized in Phytophthora and downy mildew species (Tyler et al., 2006;Haas et al., 2009;Baxter et al., 2010;Yin et al., 2017). Many studies have reported that RXLR effectors are involved in the suppression of PTI and/or ETI (Wang et al., 2011;Kong et al., 2017;Fan et al., 2018) Additionally, some of them can trigger immune response-related cell death, for example Phytophthora sojae Avh238 , Phytophthora capsici Avh1 , and Plasmopara viticola RXLR16 (Xiang et al., 2017).
After being secreted by oomycetous pathogens, the RXLR effectors are transported to a range of subcellular localizations in plant cells, including the nucleus, cytoplasm, or plasma membrane, which often associates with their functions and/or mode of action (McLellan et al., 2013;Xiang et al., 2016). The study of subcellular localization of RXLR effectors from Phytophthora infestans revealed that the nucleocytoplasmic distribution in plant cells is the most common pattern (Wang et al., 2019a). P. sojae Avh238 triggers cell death when it is present in the nucleus . In addition, plasma membrane localization of P. sojae Avh241 is required for its cell death-inducing activity (Yu et al., 2012).
Peronophythora litchii is one of the most destructive oomycete pathogens, causing downy blight on litchi fruits as well as tender leaves and panicles rot of litchi plants (Zheng et al., 2019). As a hemibiotrophic pathogen, P. litchii undergoes biotrophic and necrotrophic phases during infection. However, fewer studies have been conducted on the functions of P. litchii genes, hence there is scarcity of information about its pathogenesis and the litchi-P. litchii interaction mechanisms (Jiang et al., , 2018Kong et al., 2019). The identification and/or investigation of RXLR effectors in P. litchii lags behind that for other Phytophthora and downy mildew species, with only bioinformatics prediction of 245 putative RXLR effector genes (Ye et al., 2016). Therefore, exploring the roles of P. litchii RXLR effectors in host-pathogen interaction could potentially reveal mechanisms underlying oomycete pathogenicity and host resistance, which would be beneficial for developing disease control strategies.
In this study, systematic screening identified three P. litchii RXLR effectors, PlAvh23, PlAvh133, and PlAvh142, that are able to induce cell death by transient expression in Nicotiana benthamiana. Further experiments showed that PlAvh142 could induce cell death in different plants, therefore we focused on the investigation of PlAvh142 functions. We found that the internal repeats are indispensable for the cell death-inducing activity. PlAvh142 could localize in both cytoplasm and nucleus in the plant cell, but its cytoplasmic localization was demonstrated to be essential for triggering cell death.
Through virus-induced gene silencing (VIGS) assays, we found that cell death triggered by PlAvh142 is dependent on RAR1, SGT1, and HSP90, which suggests that PlAvh142 might be perceived by the innate immune system in plant. Finally, by genetic manipulation we proved that PlAvh142 is important for P. litchii infection to its host plant litchi. The work provides a critical foundation for further dissection of the roles of P. litchii RXLR effectors in interaction with plant immunity.

| PlAvh142 can induce cell death in plants
To systematically investigate the functions of P. litchii RXLR effectors, 212 out of 245 predicted RXLR effectors (Ye et al., 2016) were successfully cloned and then transiently expressed individually in N. benthamiana to test their cell death-inducing activity. Effector gene cloning and cell-death induction analysis are summarized in Table S1. Three RXLR effectors, PlAvh23, PlAvh133, and PlAvh142, were found to be able to induce cell death at 3-7 days post-agroinfiltration (dpa) (Figure 1a (Figure 1c). Sequence analysis indicated that PlAvh142 encodes a protein of 466 amino acids with a signal peptide from 1 to 20 amino acids. Moreover, it harbours the typical N-terminal RXLR-dEER motif (50-71 amino acids) and a potential unknown functional C-terminus ( Figure S1). Overall, we identified RXLR effectors from P. litchii that could induce plant cell death.

| Expression of PlAvh142 activates various immune responses in N. benthamiana
Cell death triggered by ETI is often considered as a part of the defence response resulting in suppression of disease progress (Balint-Kurti, 2019). In general, some other immune responses may precede this exhibition, including ROS accumulation, callose deposition, and changes in levels of phytohormones (Asai and Yoshioka, 2009;Deb et al., 2018;Bürger and Chory, 2019). Therefore, we assessed whether PlAvh142 could trigger oxidative burst production or cal- PlAvh142 is able to alter hormone signalling pathways in planta, the salicylic acid (SA)-dependent defence pathway marker genes NbPR1 and NbPR2, jasmonic acid (JA)-dependent defence pathways marker F I G U R E 1 Analysis of the cell death phenotype of RXLR effectors from Peronophythora litchii. (a) Cell death triggered by P. litchii RXLR effectors in Nicotiana benthamiana leaves. Leaves of N. benthamiana were infiltrated with Agrobacterium tumefaciens carrying pBIN::GFP-PlAvh142. Photographs were taken at 5 days post-agroinfiltration (dpa). Green fluorescent protein (GFP) and INF1 were used as negative and positive control, respectively. (b) Cell death symptoms induced by PlAvh142 in different plants. A. tumefaciens carrying PlAvh142 was infiltrated into the leaves of N. benthamiana, Solanum melongena, and Solanum lycopersicum. Photographs were taken at 5 dpa. GFP was used as negative control. (c) Confirmation of proteins accumulation. Total proteins were extracted from N. benthamiana leaves at 2 dpa. Immunoblot analyses were performed using anti-GFP (top panel) antibody. Ponceau S staining of total protein serves as loading control (bottom panel). Representative images were chosen for the results obtained from three independent experiments gene NbLOX, and ethylene (ET)-dependent defence pathways marker gene NbERF1 (Dean et al., 2005;Pieterse et al., 2012; were chosen for quantitative reverse transcription polymerase chain reaction (RT-qPCR) analysis at different time points (0, 12, 24, and 36 hr) after agroinfiltration of N. benthamiana leaves with GFP-PlAvh142 or GFP. We found that NbPR1, NbPR2, NbLOX, and NbERF1 F I G U R E 2 Transient expression of PlAvh142-activated defence responses in Nicotiana benthamiana. (a) Reactive oxygen species (ROS) accumulation and callose deposition were observed in PlAvh142-expressing plants at 36 hr post-agroinfiltration (hpa). Representative images present the results obtained from three independent experiments. (b) Quantification of the ROS using ImageJ software. Mean ± SE was derived from three independent biological repeats (n ≥ 10), and *** denotes significant differences from the green fluorescent protein (GFP) (Student's t test: p < .01). (c) Quantification of the callose deposits per microscopic field using ImageJ software. Mean ± SE was derived from three independent biological repeats (n ≥ 10), and *** denotes significant differences from the GFP (Student's t test: p < .01) (D) Relative transcript levels of defence-related genes in N. benthamiana. Leaves were harvested for RNA extraction of PlAvh142 or GFP control at 0, 12, 24, and 36 hpa. Transcript levels of candidate genes in PlAvh142-treated leaves were normalized to that of GFP control, which was arbitrarily set as 1. The constitutively expressed gene NbEF1α was used as internal reference, according to the 2 −ΔΔCt method. Data represent means ± SD from three independent biological repeats, asterisks denote significant differences from the control group (Student's t test: *p < .05; ***p < .01) were significantly up-regulated in the PlAvh142-expressing leaves at 24 and 36 hpa, compared with the GFP-expressing samples (Figure 2d), indicating an induction of phytohormone signalling by PlAvh142 in N. benthamiana. Overall, our results suggested that the expression of PlAvh142 can activate various defence responses in planta.

| The internal repeats are indispensable for PlAvh142-inducing cell death
W, Y or L motifs exist in some RXLR effectors (Win et al., 2012); however, none of them was detected in PlAvh142 (Ye et al., 2016). To further dissect the functions of PlAvh142, the conserved protein domain was analysed and predicted by the web-based program Simple Modular Architecture Research Tool (SMART, http://smart.emblheide lberg.de/). The prediction results show that in addition to the RXLR region, PlAvh142 comprises two internal repeats (IRs), IR1 (107-234 amino acids) and IR2 (324-457 amino acids), in its C-terminus (Figures 3a and S1). The alignment of IR1 and IR2 showed 25% identity and 52% similarity by BLAST analysis performed on NCBI (https ://blast.ncbi.nlm.nih.gov/Blast.cgi). To analyse the role of RXLR and IR motifs in cell death-inducing activity, four truncated PlAvh142 variants were constructed and transiently expressed in N. benthamiana

| Cytoplasmic localization is critical for PlAvh142-triggered cell death
To gain insight into the subcellular localization of PlAvh142 in the plant cell, N-terminal GFP-tagged PlAvh142 (without the signal peptide) was transiently expressed in N. benthamiana leaves, and fluorescence was observed by confocal microscopy. The free monomeric red fluorescent protein (RFP) was coexpressed with GFP-PlAvh142 and used as a marker to delineate the nucleus and cytoplasm. GFP-PlAvh142 fusion protein could be detected in both cytoplasm and nucleus of plant cells (Figure 4a). It is documented that the cell death-inducing activity of effectors is often determined by its subcellular location (Du et al., 2015). Hence, to evaluate which subcellular localization of PlAvh142 is essential for the cell death induction, we forced GFP-PlAvh142 to the nucleus or cytoplasm by attaching a nuclear localization signal (NLS) or nuclear export signal (NES), and assessed their cell death-inducing activity, respectively. Constructs with mutated NLS (nls) or NES (nes) fused to GFP-PlAvh142 were included as controls. The 2.5 | RAR1, SGT1, and HSP90 are required for PlAvh142-induced cell death in N. benthamiana ETI mediated by intracellular immune receptors usually involves a set of downstream components (Chiang and Coaker, 2015).
For example, HSP90, SGT1, RAR1, NDR1, and EDS1 are reported to be associated with signal transduction in this process (Shirasu, 2009;Bhattacharjee et al., 2011;Knepper et al., 2011). In order to determine whether these plant innate immunity components are associated with PlAvh142-induced cell death, virus-induced gene silencing was used to individually knock down these genes in N. benthamiana. Two weeks after inoculation with Agrobacterium carrying the VIGS constructs, PlAvh142 was expressed in these silenced plants and then cell death was scored 5 days later. We observed that silencing of the RAR1, SGT1, or HSP90 significantly compromised the cell death induced by PlAvh142 (Figure 5a,b).
However, the cell death proportion in NDR1-or EDS1-silenced plants was similar to that of the control plants ( Figure S2a

| PlAvh142 is up-regulated in P. litchii zoospores and during the early phase of infection
In order to investigate the biological function of PlAvh142 in P. litchii development and pathogenicity, we assessed the expression profile of PlAvh142 during different stages including mycelial growth, zoospore development, and infection of litchi leaf. The results showed that PlAvh142 was highly up-regulated in zoospores and infection stage (at 1.5, 3, 6, 12 or 24 hr post-inoculation [hpi]) in comparison to mycelia. The highest expression peak appeared at 3 hpi, and then rapidly declined ( Figure 6). However, the relative expression level at 24 hpi was still several fold higher than that of the mycelia. The accumulation of PlAvh142 transcript in zoospores and infection phase suggests that it might play a role in P. litchii infection.

| PlAvh142 contributes to P. litchii virulence
To further explore the possible role played by PlAvh142 during P. litchii infection to its native host litchi, we generated the PlAvh142 knockout mutants using the CRISPR/Cas9 gene editing system.
The single guide (sg) RNA targeted PlAvh142 were designed using the web tool EuPaGDT (http://grna.ctegd.uga.edu/) and gene replacement strategy schematically displayed in Figure 7a. Finally, we successfully generated three mutants, T14, T22, and T46, in which were also obtained and verified by RT-qPCR ( Figure S3). The mycelial growth of all mutants mentioned above was identical to that of wildtype strain SHS3 ( Figure S4).
Next, we inoculated the tender leaves of litchi plants with 100 zoospores from the wild-type strain, knockout mutants, or overexpressed mutants. An unsuccessful knockout transformant T37 was included, serving as negative control. We observed that the PlAvh142 knockout mutants caused fewer disease symptoms in litchi leaves at 48 hpi compared to that of the wild-type strain (Figure 7c,d). In contrast, the overexpressed mutants caused more severe disease symptoms (Figure 7c,d). These results indicate that PlAvh142 contribute to the P. litchii virulence during infection in its native host litchi.

| D ISCUSS I ON
Like other biotrophic and hemibiotrophic oomycete or fungal pathogens, P. litchii must overcome plant defences to establish host colonization. Pathogen effectors play important roles in subverting plant immunity. In oomycetes, much attention has been focused on identification and functional analysis of RXLR effectors over the past decade (Wang et al., 2011(Wang et al., , 2019Xiong et al., 2014;Huang et al., 2019).
Although the genome sequence of P. litchii has been published with 245 RXLR predicted (Ye et al., 2016), the function of these effectors during P. litchii infection remains unknown. In this study, we used the N. benthamiana model system for systematic screening of P. litchii RXLR effectors with plant cell death-inducing activity. Here we reported that three effectors identified were able to trigger cell death in N. benthamiana, among which PlAvh142 could trigger plant cell death in a broad spectrum of plants and therefore it was chosen for further investigation. Unfortunately, we failed to test whether PlAvh142 could trigger cell death in its host plant, as we were not successful in a particle bombardment assay on litchi leaves despite repeated attempts.
Bioinformatic analysis showed that about 15% of the 245 predicted P. litchii RXLR effectors harbour IRs (Table S1) MY, P. litchii mycelia grown in CJA medium; ZO, zoospores. Litchi leaves inoculated with P. litchii zoospores were harvested at 1.5, 3, 6, 12, and 24 hr post-inoculation (hpi). The relative expression level was calibrated to the levels for the MY set as 1. The constitutively expressed gene PlActin was used as internal reference. Error bars represent the SD of three biological replicates prokaryotes, and are considered to be involved in protein-protein interaction (Andrade et al., 2001;Pawson and Nash, 2003;Björklund et al., 2006). A recent report showed that P. sojae RXLR effector PsAvh23 contains two IRs and at least one IR that is required for its interaction with host target protein . In this study we proved that both IRs in PlAvh142 were required for its cell death-inducing activity. To our knowledge, this is the first report on the requirement of IRs for RXLR inducing plant cell death. However, it remains unclear whether these two IRs are required for interaction (if any) between PlAvh142 and its target protein(s) in the plant cells. RXLR effectors could localize in different compartments of the host cells, in correspondence to their various molecular/cellular functions during the host-pathogen interaction (Liu et al., 2018;Wang et al., 2019a). Thus, perception of effector proteins by the cognate receptor(s) is frequently associated with their localized position, for example activation of R1-mediated HR and resistance required localization of the R1/AVR1 pair in the nucleus, although both AVR1 and its cognate R protein R1 could be observed in cytoplasm and nucleus (Du et al., 2015).
In this study we found that localization of PlAvh142 in the cytoplasm of the plant cell was sufficient and essential for inducing cell death, which is different from many other identified cell death-inducing RXLR effectors shown to be localized to the plasma membrane or nucleus in plant cells (Yu et al., 2012;Asai et al., 2018;Yin et al., 2019). A weak cell death was observed in plant cells expressing the nuclear localized PlAvh142 variant, which may be due to a small amount of protein residue in the cytoplasm. We infer that the cell death triggered by PlAvh142 effector may depend on its recognition in the cytoplasm of the plant cell.
A conserved chaperone complex consisting of HSP90, SGTI, and RAR1 is known to stabilize and sustain NLR-mediated ETI responses (Azevedo et al., 2002;Shirasu, 2009), and is required for plant cell death triggered by PvAvh74 (Yin et al., 2019). SGT1 and HSP90, rather than RAR1, are required for PpE4-triggered cell death (Huang et al., 2019). In the cases of P. infestans AVR-blb2 and PITG_22798, cell death-inducing activity is dependent on SGT1 (Oha et al., 2014;Wang et al., 2017). Besides, SGT1 is also involved in PTI and plant defence against viruses (Huitema et al., 2005;Boter et al., 2007). In the present study, silencing of (c) Pathogenicity assays of WT and PlAvh142 mutants on litchi leaves. Tender leaves were inoculated with 100 zoospores from WT, T37 (negative control), three PlAvh142 knockout mutants T14, T22, and T46, and two overexpression mutants OE7 and OE10. Disease symptoms were visually monitored over a period of 48 hr and photographs were taken at 48 hr post-inoculation (hpi). (d) Lesion length was measured after 48 hpi. Different letters represent significant differences (p < .01; Duncan's multiple range test). Bars represent medians and boxes the 25th and 75th percentiles. There were 30 leaves used in each of the three biological replicates underlie this effector. We speculate that either PlAvh142 targets a conserved and critical plant protein guarded by NLR genes or direct NLR recognition is conserved in various plants. Despite this, we cannot rule out the possibility that cell death triggered by PlAvh142 is mediated by an unknown mechanism.
Hemibiotrophic pathogens need to keep the host cell alive before establishing their colonization in the biotrophic stage, and later trigger cell death to promote the necrotrophic infection (Qutob et al., 2002). Some apoplastic or cytoplasmic effectors from Phytophthora pathogens display elicitor activity, that is, they could trigger plant immunity responses, and concurrently contribute to the virulence or promote pathogen colonization. Examples include PsXEG1 (Ma et al., 2015(Ma et al., , 2017Wang et al., 2018b), Avh238 , and PpE4 (Huang et al., 2019). Another similar example was reported in the cell death-inducing RXLR effector, PcAvh1, which is a virulence factor of P. capsici as its deletion mutants displayed reduced pathogenicity in contrast to the more aggressive overexpression mutants . Such a seemingly contradictory phenomenon was also observed for PlAvh142 in this study, and we raise the following hypotheses in an attempt to explain this. First, the accumulation of PlAvh142 protein during P. litchii infection to the native host plant may be insufficient to trigger cell death and/ or immune responses under natural conditions. Second, cell death triggered by one RXLR effector could be suppressed by the other cooperative effectors during pathogen infection, which has been already reported (Wang et al., 2011). Therefore, PlAvh142 is still able to enhance colonization and execute its virulence function when its elicitor activity is blocked. Alternatively, the cell death induced by PlAvh142 may contribute to the transition from biotrophy to necrotrophy and thus positively regulate P. litchii virulence. Although it is generally accepted that RXLR effectors facilitate pathogen infection mainly by modulating plant immune system, the functional relationship (if any) between PlAvh142's cell death-inducing activity and contribution to virulence awaits elucidation; alternatively, a mechanism(s) other than cell death induction underlying PlAvh142 virulence function needs to be uncovered.
Overall, we report here, for the first time, that an RXLR effector secreted by P. litchii acts as an elicitor that triggers immune responses in plants. The possible mechanism involved in perception of PlAvh142 will be very useful for exploring the potential resistance genes or materials for the litchi plant, which provides insights into novel disease control strategies. The next step is to identify the potential PlAvh142-interacting protein(s) to reveal the functions of RXLR effector in litchi-P. litchii interaction, for a better understanding of the biological functions of RXLR effectors.

| Plasmid construction
All the primers used in this study are listed in Table S2. The PCR fragments were amplified by Phanta Max Super-Fidelity DNA Polymerase

| RNA extraction, cDNA synthesis, and expression analysis of PlAvh142
Mycelia and litchi leaves infected with zoospores suspension of P. litchii were harvested at the indicated time points and RNA was extracted using All-In-One DNA/RNA Mini-preps Kit (Bio Basic) according to the recommended protocol. All cDNAs were synthe-

| Agroinfection assay in N. benthamiana
The PlAvh142 gene was amplified from P. litchii cDNA and then cloned into the PVX vector pGR107 and pBINGFP2, respectively.

| Callose and ROS staining
To observe callose deposition and ROS accumulation in planta, the whole leaves of N. benthamiana were harvested at 36 hpa. For callose deposition assay, leaves were stained with 0.01% aniline blue in 150 mM K 2 HPO 4 buffer 1-2 hr after destaining in 96% ethanol (Sohn et al., 2007) and subsequently imaged by Olympus BX53 microscopy system. For ROS accumulation assays, leaves were visualized using diaminobenzidine-HCl solution (1 mg/ml, pH 3.8) in darkness for 8-12 hr and subsequently destained with 96% ethanol . The quantification of callose deposition and ROS accumulation was calculated using ImageJ software. At least three leaves were tested in each independent experiment. The experiments were repeated at least three times.

| Protein extraction and western blot analysis
The leaves of N. benthamiana that were infiltrated with A. tumefaciens were ground into powder in liquid nitrogen and vigorously mixed with 0.5 ml of precooled radioimmunoprecipitation assay buffer (RIPA buffer) (250 mM Tris-HCl [pH 7.6], 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% sodium dodecyl sulphate [SDS] [Thermo Fisher Scientific]). After 5 min of incubation on ice, the samples were centrifuged at 14,000 × g for 15 min to obtain the supernatant. After adding loading buffer and boiling for 5 min, total proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) gels. Then the proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad) followed by blocking in 5% non-fat milk dissolved in PBST (phosphate-buffered saline + 0.1% Tween 20). Mouse anti-GFP monoclonal antibody (Abbkine) was used at 1:5,000 dilution to detect the corresponding fusion proteins. The membranes were washed three times with PBST and incubated with a goat anti-mouse antibody (1:10,000) (Mei5 Biotechnology). Proteins were visualized by Efficient Chemiluminescence kit (Genview), and photographs were taken under the imaging system (Tanon).

| Confocal microscopy
For fluorescence observations, patches of N. benthamiana leaves were cut after 2 dpa and used for confocal imaging on a Nikon A1 laser scanning microscope with a 40× objective lens. RFP or GFP fluorescence was observed at an excitation wavelength of 561 or 488 nm, respectively.

| Pathogenicity assays
For pathogenicity assays, zoospores were inoculated on the tender leaves of litchi (Guiwei), which were collected from the litchi orchard in South China Agricultural University, Guangzhou, Guangdong province. One hundred zoospores of each strain were inoculated on the center of the tender leaf, and kept at 80% humidity in 12 hr light/12 hr darkness at 25 °C. Each strain was tested on no fewer than 10 leaves. The symptoms were observed and the lesion diameter was measured at 48 hpi. The experiments were repeated at least three times.

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.