Functional Identification of Novel Cell Death-inducing Effector Proteins from Magnaporthe oryzae

Background Secreted effector proteins play critical roles in plant-fungal interactions. The Magnaporthe oryzae genome encodes a large number of secreted proteins. However, the function of majority of M. oryzae secreted proteins remain to be characterized. We previously identified 851 in planta-expressed M. oryzae genes encoding putative secreted proteins, and characterized five M. oryzae cell death–inducing proteins MoCDIP1 to MoCDIP5. In the present study, we expand our work on identification of novel MoCDIP proteins. Results We performed transient expression assay of 98 more in planta-expressed M. oryzae putative secreted protein genes, and identified eight novel proteins, MoCDIP6 to MoCDIP13, that induced plant cell death. Yeast secretion assay and truncation expression analysis revealed that the signal peptides that led the secretion of proteins to the extracellular space, were required for cell death inducing activity of the novel MoCDIPs except for MoCDIP8. Exogenous treatment of rice seedlings with recombinant MoCDIP6 or MoCDIP7 resulted in enhanced resistance to blast fungus, indicating that the two MoCDIPs trigger cell death and elicit defense responses in rice. Conclusions The newly identified MoCDIP6 to MoCDIP13, together with previously identified MoCDIP1 to MoCDIP5, provide valuable targets for further dissection of the molecular mechanisms underlying the rice-blast fungus interaction. Electronic supplementary material The online version of this article (10.1186/s12284-019-0312-z) contains supplementary material, which is available to authorized users.


Background
Plant pathogenic fungi are the causal agents of many of the world's most destructive plant diseases, causing serious agricultural losses worldwide. Plant pathogenic fungi have diverse lifestyles and interact with host plants in various ways: biotrophic fungi colonize and obtain nutrients from living host tissue; necrotrophic fungi infect host tissue and harvest nutrients from dead host cells; whereas hemibiotrophic fungi combine an initial biotrophic phase with a subsequent necrotrophic phage (Lo Presti et al. 2015). Despite the diversity of interaction manners, all plant pathogenic fungi secrete extracellular proteins to facilitate infection. These secreted proteins may function in the apoplast as virulence factors, toxins, and degradative enzymes, or within the plant cytoplasm to manipulate host cell physiology and suppress host immune response (Giraldo and Valent 2013;Kim et al. 2016). In turn, plants have evolved sophisticated immune systems to protect themselves from pathogen invasion, including a basal defense system that recognizes conserved pathogen-associated molecular patterns (PAMPs), and a second layer of immunity through the recognition of secreted effector proteins for triggering defense responses (Jones and Dangl 2006).
In a previous study, we performed transcriptome analyses on blast-infected rice leaves, and identified 851 in planta-expressed M. oryzae genes encoding putative secreted proteins. We performed transient expression of 42 in planta-expressed putative secreted protein genes, and identified five M. oryzae cell death-inducing proteins (MoCDIP1 to MoCDIP5), that induced cell death in plant cells (Chen et al. 2013). In the present study, we expand our work on identification of novel MoCDIP proteins. We cloned 98 more in planta-expressed M. oryzae putative secreted protein genes, and identified eight novel putative proteins, MoCDIP6 to MoCDIP13, that induced cell death in plant cells. We further demonstrated that exogenous treatment of rice seedlings with recombinant MoCDIP6 and MoCDIP7 resulted in enhanced resistance to rice blast, indicating that the two MoCDIP proteins trigger cell death and elicit defense responses in rice.

Identification of eight M. oryzae cell death-inducing proteins
We previously identified five in planta-expressed M. oryza secreted proteins, MoCDIP1 to MoCDIP5, as cell deathinducing proteins (Chen et al. 2013). In this study, to get more insight into the function of in planta-expressed M. oryzae secreted proteins, we conducted functional characterization of more genes from the previously identified 851-gene list. A total of 98 M. oryzae putative secreted protein genes were amplified by PCR (Additional file 1: Table S1) and were cloned into the plant expression vector pGD (Goodin et al. 2002). The M. oryzae putative secreted protein genes were investigated through agro-infiltration in leaves of Nicotiana benthamiana. Among the 98 gene constructs, eight were found to induce cell death in N. benthamiana leaves (Fig. 1a). 3,3′-Diaminobenzidine tetrachloride (DAB) staining further showed the accumulation of H 2 O 2 in agro-infiltrated N. benthamiana leaves which was correlated with cell death lesions (Fig. 1b). The severity of cell death in N. benthamiana leaves induced by the eight constructs varied. While MGG_13283 and MGG_14371 induced clear necrosis symptoms at 3 to 4 days after infiltration, the cell death phenotypes induced by MGG_01532, MGG_03354, MGG_05038, MGG_08411, MGG_12275, and MGG_ 12521, were visible at 4 to 6 days after infiltration, with relatively weak symptoms. The eight putative proteins, MGG_01532, MGG_03354, MGG_05038, MGG_08411, MGG_12275, MGG_12521, MGG_13283, and MGG_ 14371 were thus referred to as MoCDIP6 to MoCDIP13, respectively (Additional file 1: Table S1).

MoCDIP6 to MoCDIP13 genes are expressed during infection
The eight MoCDIP genes, MoCDIP6 to MoCDIP13, were screened from transcriptome profiles of M. oryzae-infected rice leaves (Chen et al. 2013). To further confirm the in planta expression pattern of the eight MoCDIP genes, a rice cultivar MH3301 was artificially inoculated with a virulent M. oryzae isolate 501-3, and quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) was performed to detect the expression of the eight MoCDIP genes in inoculated rice leaves. The results showed that MoCDIP6 to MoCDIP13 were expressed at low levels in mycelia. In contrast, the eight MoCDIP genes were all expressed at significantly higher levels in M. oryzae-inoculated rice leaves (Fig. 2). While relatively higher level of the MoCDIP7 transcript was detected in inoculated rice leaves at 24 h post-inoculation (hpi), higher expression of the rest MoCDIPs were detected in inoculated rice leaves at 48 and/or 72 hpi (Fig. 2). Overall, qRT-PCR results confirmed that the eight MoCDIP genes were expressed during infection stages.
Signal peptides of seven MoCDIPs are required for cell death-inducing activity in N. benthamiana The eight MoCDIP proteins, MoCDIP6 to MoCDIP13, are diverse in their sequences. Sequence analysis revealed that MoCDIP8 belonged to a family of endonuclease/exonuclease/phosphatase; MoCDIP10 was a putative ferritin-like family protein; and MoCDIP11 contained a conserved CFEM domain, which was unique to fungi (Kulkarni et al. 2003;Kou et al. 2017). As for MoCDIP6, MoCDIP7, MoCDIP9, MoCDIP12, and MoCDIP13, no putative conserved domains or motifs were identified (Fig. 3a). Unexpectedly, while MoCDIP6, MoCDIP7, MoCDIP9, MoCDIP10, MoCDIP11, MoC-DIP12, and MoCDIP13 were predicted to contain a signal peptide, no signal peptide was predicted in MoCDIP8 based on SignalP 3.0 (http://www.cbs.dtu.dk/ services/SignalP-3.0/) or SignalP 4.1 analysis (http:// www.cbs.dtu.dk/services/SignalP-4.1/). A yeast secretion system was exploited to experimentally corroborate the predicted signal peptides. The sequences of full-length and truncated version lacking the predicted signal peptide sequence of MoCDIPs (referred to as FL-MoCDIPs and NS-MoCDIPs, respectively) were fused in frame with a truncated yeast invertase Suc2 gene lacking its original signal peptide sequence. The fusion constructs were transformed into an invertase-deficient yeast mutant DBYα2445 (Lee et al. 2006). The transformed yeasts were plated on a sucrose selection medium. Consistent with the prediction, transformation of the fusion con-  b Functional validation of the predicted signal peptides of seven MoCDIPs based on a yeast secretion assay system. pYST-2, yeast colony transformed with an empty vector containing a truncated yeast invertase Suc2 gene lacking its original signal peptide sequence. FL-MoCDIPs-Suc2, yeast colonies transformed with full length MoCDIPs fused in frame with Suc2. NS-MoCDIPs-Suc2, truncated version lacking the predicted signal peptide sequence of MoCDIPs fused in-frame with Suc2. Yeast colonies grown on SD/−Leu medium were replica-plated onto sucrose selection medium for about 3 days. Strains unable to secrete invertase were inhibited on sucrose medium (white color), whereas strains able to secrete invertase could grow on sucrose medium (red color). MoCDIP13 were also cloned into the expression vector pGD for agro-infiltration assay. The results showed that, while transient expression of FL-MoCDIPs repeatably induced cell death in N. benthamiana leaves, agro-infiltration of NS-MoCDIPs failed to induce cell death (Fig. 3c). Consistently, no H 2 O 2 accumulation was observed in N. benthamiana leaves infiltrated with Agrobacterium tumefaciens carrying the NS-MoCDIPs constructs (Fig. 3c). RT-PCR analyses were carried out to examine the expression of MoCDIPs in the infiltrated N. benthamiana leaves, and the results showed that both the FL-MoCDIPs and NS-MoCDIPs were expressed at similar mRNA levels (Fig. 3d). Taken together, the results suggested that the signal peptides of MoCDIP6, MoCDIP7, MoCDIP9, MoCDIP10, MoCDIP11, MoCDIP12, and MoCDIP13 that led the secretion of the proteins to the extracellular space, were required for cell death inducing activity.  (Fig. 4a). The recombinant MoCDIP6 and MoCDIP7 solutions (10 μM) were sprayed onto rice seedlings. At two days after spray, rice seedlings displayed wilt and their leaves showed necrosis symptoms (Fig. 4b, Additional file 2: Figure S1). In contrast, rice seedlings treated with control buffer or MBP solution had no visible wilt or necrosis symptoms. These results confirmed cell death inducing activity of MoCDIP6 and MoCDIP7.
To determine whether MoCDIP6 and MoCDIP7 can elicit defense responses in rice, we sprayed the recombinant MoCDIP6 and MoCDIP7 solutions with a lower concentration (2 μM) onto rice seedlings, and evaluated the expression levels of four rice pathogenesis-related (PR) genes, OsCht1, OsCht3, OsNac4 and OsPR1b. When sprayed with MoCIDP6 or MoCDIP7 at 2 μM, rice plants did not exhibit wilt or necrosis symptoms. Both MoCIDP6 and MoCDIP7 induced the expression of the four PR genes (Fig. 5). After 24 h pre-treatment with the lower dosage level of MoCIDP6 and MoCDIP7, the rice seedlings were further inoculated with the virulent fungal isolate Guy11. At 5 days post-inoculation (dpi), MoCIDP6-and MoCDIP7-pretreated plants, and control rice plants pretreated with the mock buffer all exhibited blast disease symptoms. However, while control plants exhibited typical and severe blast disease symptoms, most lesions observed on MoCIDP6-or MoCDIP7-pretreated plants were small and constrained (Fig. 6a). Consistently, disease lesion areas on leaves of MoCIDP6-or MoCDIP7-pretreated plants were significantly smaller than that on control plants (Fig. 6b). These results, therefore, demonstrated that MoCDIP6 and MoCDIP7 elicited defense responses in rice.

MoCDIPs are dispensable for pathogenicity
To study whether MoCDIPs were required for pathogenicity of M. oryzae, we performed overexpression and disruption analyses of the eight MoCDIP genes. We successfully generated overexpression mutants (OE) for all eight MoCDIP genes and knock out mutants (Δ) for MoCDIP6, MoCDIP7, MoCDIP8, MoCDIP10, and MoC-DIP11 in Guy11 background (Table 1). The OEor Δmutants and the wild type isolate Guy11 were characterized for their vegetative growth rate, sporulation, and pathogenicity. Compared with the wild type Guy11, OE-MoCDIP6 showed a slower vegetative growth rate on oat medium, whereas the other OEor Δ-mutants exhibited similar growth rate to that of Guy11 (Table 1). There was no significant difference in sporulation between the OEor Δ-mutants and the wild type Guy11 (Table 1). The pathogenicity of the OEor Δ-mutants was determined based on infection assays on detached barley leaves and on rice seedlings. Infection assays on detached barley leaves showed that the OEor Δ-mutants developed similar disease lesions compared to that of Guy11. In addition, the OEor Δ-mutants were highly virulent to rice cv. Nipponbare. No significant difference was observed for disease severity between the OEor Δmutants and Guy11 (Table 1). These results suggested that the MoCDIP genes, at least MoCDIP6, MoCDIP7, MoCDIP8, MoCDIP10, and MoCDIP11 were dispensable for pathogenicity of the rice blast fungus.

Discussion
Secreted effector proteins play critical roles in interactions between plants and phytopathogenic fungi (Giraldo and Valent 2013). The M. oryzae genome encodes a large number of secreted proteins (Dean et al. 2005;Yoshida et al. 2009;Choi et al. 2010). However, the function of majority of M. oryzae secreted proteins remain unknow. In our previous study, we performed transient expression assay of in planta-expressed M. oryzae secreted protein genes to screen cell death-inducing effector proteins (Chen et al. 2013). However, we had difficulty cloning a large number of in planta-expressed M. oryzae ORFs from cDNA generated from blast-infected rice leaves, mainly because the proportion of fungal RNA was relatively low (Chen et al. 2013). In the present study, we performed PCR or overlapping PCR using M. oryzae genomic DNA as template to clone those M. oryzae genes without introns or with only one intron. In addition, cDNA generated from mRNA of M. oryzae cultured under nitrogen starved conditions (Sharpee et al. 2017) was used as template for amplification of M. oryzae genes. We successfully cloned 98 more in planta-expressed M. oryzae putative secreted protein genes in total. The newly cloned genes, together with the 42 previously cloned genes (Chen et al. 2013), could also serve as resources for identifying M. oryzae secreted protein effectors with ability to suppress plant cell death (Sharpee et al. 2017), or having other different functions.
Studies from diverse fungal pathogens have shown that many fungal secreted protein effectors possessed abilities to induce plant cell death (Chen et al. 2013;Fang et al. 2016;Anderson et al. 2017). For example, using the agro-infiltration MoCDIP11, MoCDIP12, and MoCDIP13 were predicted to contain a N-terminal signal peptide, and the function of the signal peptides for directing proteins to the secretory pathway was experimentally validated in yeast (Fig. 3b). Furthermore, truncation analysis revealed that the signal peptides were required for the seven proteins expressed in N. benthamiana leaves to induce cell death, suggesting that the secretion characteristics was required for cell death inducing activity of the seven MoCDIPs. Our results, together with those previously reported, demonstrated that secreted proteins from M. oryzae are enriched for cell death-inducing effectors.
Cell death plays a central role in plant-pathogen interactions (Kanneganti et al. 2006). As for hemibiotrophic and necrotrophic pathogens, some of cell death-inducing secreted protein effectors were believed to have function in facilitating the colonization during the necrotrophic phase (Sharpee et al. 2017). Indeed, many of these cell death-inducing protein effectors have been demonstrated to be required for pathogenicity, or contribute to the virulence of the pathogens (Ma et al. 2015;Qutob et al. 2006). On the other hand, some cell death-inducing proteins were shown to be dispensable for virulence of pathogens. For example, disruption of the MSP1 gene , and the MoNLP family genes (Fang et al. 2017) in M. oryzae did not impair the fungal virulence on rice. In the present study, we also observed that overexpression of the eight identified MoCDIP genes, or targeted deletion of MoCDIP6, MoCDIP7, MoCDIP8, MoCDIP10, or MoC-DIP11 did not affect virulence of M. oryzae isolate Guy11 (Table 1). Since there were a number of cell death inducing secreted proteins in M. oryzae, one of the reasons for no virulence changes in the knock out mutants could be due to functional redundancy of closely related secreted proteins. Hence, further investigations would be needed to determine the biological roles of these cell death-inducing proteins in the interaction between M. oryzae and rice. Many cell death-inducing proteins have been shown to elicit plant immune responses. Recent studies have revealed several cell death-inducing proteins that are recognized as PAMPs. For instance, multiple cytotoxic NLPs have been found to harbor a pattern of 20 amino acid residues (nlp20) that is recognized by Brassicaceae plant species (Böhm et al. 2014). The nlp20 motif interacts with the Arabidopsis LRR receptor protein RLP23, leading to PAMP-triggered immunity (PTI) transmitted via the RLP23-SOBIR1-BAK1 complex (Albert et al. 2015). XEG1, a Phytophthora sojae glycoside hydrolase 12 protein with cell death-inducing activity, and Rhynchosporium commune Cell Death Inducing 1 (RcCDI1) were also identified as PAMPs (Ma et al. 2015;Franco-Orozco et al. 2017). In M. oryzae, MoHrip1, MoHrip2, and MSP1 were demonstrated to be elicitors of defense responses in rice (Chen et al. 2012;Chen et al. 2014;Wang et al. 2016). In this study, we were able to express and purify soluble recombinant MoCDIP6 and MoCDIP7 from E. coli. We observed that application of ectopically expressed MoCDIP6 and MoCDIP7 induced the expression of PR genes in rice seedlings and increased the resistance of rice to M. oryzae, indicating that the two proteins triggered both cell death and immune responses in rice. It remains to be determined whether the other MoCDIPs can elicit plant immune responses and how plant cells recognize and respond to these proteins.
In summary, we identified eight novel M. oryzae cell death-inducing proteins, MoCDIP6 to MoCDIP13. Exogenous treatment of rice seedlings with recombinant MoCDIP6 and MoCDIP7 resulted in enhanced resistance to blast fungus, indicating that the two MoCDIP Disease severity was scored by following 0-5 scale (0-1: Resistant, 2: moderately resistant, 3: moderately susceptible, 4-5: severe susceptible); Statistical significances (*P < 0.05 or **P < 0.01) between the MoCDIPs mutants and the wild type Guy11 were analyzed by student's t-test; ND Not determined proteins triggered cell death and elicit defense responses in rice. The newly identified MoCDIP6 to MoCDIP13, together with previously identified MoCDIP1 to MoC-DIP5, provide candidate targets for further studies to better understand the molecular mechanisms underlying the rice-M. oryzae interaction.

Plant materials and fungal isolates
Rice (Oryza sativa) materials used in this study were Nipponbare and MH3301. M. oryzae isolates used in this study included 70-15, Guy11 and 501-3.
Cloning of M. oryzae genes encoding putative secreted proteins M. oryzae genes encoding putative secreted proteins selected from the previously identified 851-gene list (Chen et al. 2013) were cloned by PCR amplification using gene-specific primers (Additional file 3: Table S2). Genomic DNA of M. oryzae isolate 70-15 was used as template for amplification of those M. oryzae genes without intron or with only one intron by PCR or overlapping PCR, and cDNA generated from mRNA of 70-15 cultured in minimal media lacking nitrogen was used as template for amplification of M. oryzae genes with introns. The amplified genes were cloned into a modified plant expression vector pGD (Goodin et al. 2002) that allowed for TA cloning of PCR products (Chen et al. 2009). All constructs were confirmed by sequencing.

Agro-infiltration assays and DAB staining
The pGD-based M. oryzae gene constructs were transformed into A. tumefaciens strain GV3101 through electroporation. The A. tumefaciens clones containing M. oryzae gene constructs were cultured in liquid YEP media supplemented with rifampicin (50 μg·mL − 1 ) and kanamycin (50 μg·mL − 1 ). The A. tumefaciens cultures were collected by centrifugation, washed with sterile double-distilled H 2 O, and then resuspended in agro-infiltration buffer (10 mM MES, 10 mM MgCl 2 , and 150 μM actosyringone, pH 5.7) at an OD600 of 0.5 at room temperature for 2 h. Agro-infiltration experiments were carried out on leaves of 6-week-old N. benthamiana plants using needleless syringes.
For DAB staining, N. benthamiana leaves were incubated in DAB solution (1 mg·mL − 1 ) for 8 h at 25°C in the dark. After staining, the leaves were washed with sterile double-distilled H 2 O, and were then boiled in bleaching solution containing ethanol and acetic acid (3:1) for 10-15 min.

Gene expression analysis
Total RNA was extracted from M. oryzae mycelia, or rice leaves by using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA) and treated with RNasefree DNase I (Ambion, Austin, TX). One microgram of total RNA was used for reverse transcription by using the Promega Reverse Transcription System (Promega, Madison, WI). Transcript levels of MoCDIPs and rice PR genes were determined by qRT-PCR under standard conditions with gene-specific primers (Additional file 3: Table S2). The experiments were performed with three replications.

Yeast secretion assays
The pYST-2 vector (Lee et al. 2006), which contains a truncated invertase gene (Suc2) lacking the coding sequence of signal peptide, was used for validation of predicted signal peptides of the identified MoCDIPs. PCR fragments of FL-MoCDIPs and NS-MoCDIPs without the stop codon were amplified using specific primers (Additional file 3: Table S2). The fragments were cloned into pYST-2 using the restriction enzyme digestion and ligation method to fuse in-frame with Suc2. The resulting constructs were transformed into the invertase-deficient yeast mutant DBYα2445 (Lee et al. 2006). Transformants were cultured on SD/−Leu medium (1.7% yeast N base, 2% glucose, 0.5% ammonium sulfate, 0.069% complete amino acid supplement mixture minus leucine, and 2% agar), and yeast colonies were replica-plated onto sucrose selection medium (1% yeast extract, 2% peptone, 2% sucrose, and 2% agar) for secretion assays.

Expression and purification of recombinant proteins
DNA fragments of NS-MoCDIP6, NS-MoCDIP7, NS-MoCDIP9, NS-MoCDIP10, NS-MoCDIP11, NS-MoC-DIP12, or NS-MoCDIP13 digested with BamHI from the pGD-based gene constructs were inserted into the BamHI site of the pMAL-c2× vector (New England Biolabs, Ipswich, USA), which harbors a MBP tag at the Nterminus of the fusion proteins. The resulting constructs were transformed into E. coli strain BL21. Expression of the recombinant MBP-NS-MoCDIPs were induced by 0.5 mM IPTG at 16°C for overnight. MBP fusion proteins were affinity purified with amylose resin (New England Biolabs) according to manufacturer's instructions. The recombinant proteins were evaluated by SDS-PAGE and staining with Coomassie Blue. The protein concentration was determined by using the Bradford protein assay kit (Bio-Rad Laboratories, Hercules, CA).

Generation of overexpression and Knock-out mutants of MoCDIPs, and phenotype analyses
To overexpress the novel identified MoCDIPs, DNA fragments of FL-MoCDIPs digested with BamHI from the pGD-based gene constructs were inserted into the BamHI site of the pSM56 vector (Bourett et al. 2002), which contained a RP27 promoter for strong constitutive expression in M. oryzae. The resulting constructs were transformed into protoplasts of Guy11 through PEGmediated transformation (Sweigard et al. 1992). Colonies growing on selection media with carbenicillin (50 μg·mL − 1 ) and hygromycin (150 μg·mL − 1 ) were evaluated by PCR to confirm the presence of the transgenes, and overexpression mutants were screened by qRT-PCR using specific primers (Additional file 3: Table S2).
To generate disruption mutants, genomic DNA fragments about 0.8 kb in length flanking the coding region of the targeted MoCDIP genes were amplified and were cloned into the pCX62 vector to flank the hygromycin phosphotransferase (hph) gene cassette. The resulting constructs were used for transformation of protoplasts of Guy11. Hygromycin-resistant colonies were screened by PCR for disruption of the MoCDIP genes.
Growth of fungal mycelia was tested by placing mycelium blocks (5 mm in diameter) on the center of a CM medium plate. The inoculated plates were incubated at 28°C under dark conditions. The diameter of colonies was measured at 10 days. For quantification of sporulation, mycelium blocks (5 mm in diameter) were cultured on rice bran medium at 28°C under dark conditions for 5 days, then were incubated at 28°C under light for 3 days to induce spores. The plates were washed with 3 mL of distilled water. The conidial suspension was filtrated through a 100 μm mesh, and the number of spores was counted by hemocytometer.

Plant inoculations
Infection assays on rice seedlings were performed in a greenhouse following a spraying inoculation method as previously described (Tian et al. 2016). Two-week-old rice seedlings were spray-inoculated with M. oryzae spores at a concentration of 2 × 10 5 conidia·mL − 1 . The inoculated seedlings were maintained under high humidity in the dark for 24 h, and were grown at 25°C, under high humidity with a 16-h photoperiod for further 4-5 days.
Infection assays with detached barley leaves were conducted as previously described (Liu et al. 2011). Detached leaves of about eight-day-old barley seedlings were placed in plates with moist filter paper for drop inoculation. The plates were placed in dark at 25°C for 24 h, and were then incubated at 25°C with a 16-h photoperiod for further 4 days.