The MoPah1 phosphatidate phosphatase is involved in lipid metabolism, development, and pathogenesis in Magnaporthe oryzae

Abstract

As with the majority of the hemibiotrophic fungal pathogens, the rice blast fungus Magnaporthe oryzae uses highly specialized infection structures called appressoria for plant penetration. Appressoria differentiated from germ tubes rely on enormous turgor pressure to directly penetrate the plant cell, in which process lipid metabolism plays a critical role. In this study, we characterized the MoPAH1 gene in M. oryzae, encoding a putative highly conserved phosphatidate phosphatase. The expression of MoPAH1 was up‐regulated during plant infection. The MoPah1 protein is expressed at all developmental and infection stages, and is localized to the cytoplasm. Disruption of MoPAH1 causes pleiotropic defects in vegetative growth, sporulation, and heat tolerance. The lipid profile is significantly altered in the Mopah1 mutant. Lipidomics assays showed that the level of phosphatidic acid (PA) was increased in the mutant, which had reduced levels of diacylglycerol and triacylglycerol. Using a PA biosensor, we showed that the increased level of PA in the Mopah1 mutant was primarily accumulated in the vacuole. The Mopah1 mutant was blocked in both conidiation and the formation of appressorium‐like structures at hyphal tips. It was nonpathogenic and failed to cause any blast lesions on rice and barley seedlings. RNA sequencing analysis revealed that MoPah1 regulates the expression of transcription factors critical for various developmental and infection‐related processes. The Mopah1 mutant was reduced in the expression and phosphorylation of Pmk1 MAP kinase and delayed in autophagy. Our study demonstrates that MoPah1 is necessary for lipid metabolism, fungal development, and pathogenicity in M. oryzae.

MoPAH1 is involved in lipid metabolism, vegetative growth, development, MAPK signalling pathways, autophagy, and pathogenicity in the rice blast fungus.

1. INTRODUCTION

Magnaporthe oryzae is the causative agent of the rice blast disease that causes annual losses that are sufficient to feed more than 60 million people worldwide, seriously threatening global food security (Dean et al., 2012; Fernandez & Orth, 2018). As a typical filamentous ascomycete, M. oryzae uses infection‐specific structures called appressoria for plant infection. Because of the critical role of appressoria in plant penetration, appressorium formation has been extensively studied (Kong et al., 2013; Li et al., 2012). Similar structures called appressorium‐like structures (ALSs), which differentiate from hyphal tips, can also penetrate cell walls (Kong et al., 2013). The process of appressorium or ALS formation is coordinately regulated by multiple signalling pathways, including the cyclic AMP‐protein kinase A (cAMP‐PKA), and the Pmk1 and Mps1 mitogen‐activated protein kinase (MAPK) pathways, with Pmk1 playing the most critical role (Kronstad et al., 1998; Zhang et al., 2021). Additionally, autophagy regulates nutrition recycling and is involved in virulence in M. oryzae. Multiple mutants that lack key autophagy components are nonpathogenic (He et al., 2013; Kershaw & Talbot, 2009).

Enormous turgor pressure is generated in the appressorium from accumulating glycerol and facilitates the penetration peg to directly puncture the plant cell wall (Howard et al., 1991; Wang et al., 2007). The above process is energy‐consuming, and glycogen stored in conidia is rapidly metabolized to generate sufficient energy during germination (Thines et al., 2000). Mutants lacking glycogen metabolic genes AGL1 or GPH1 are able to form appressoria but are reduced in virulence on rice (Badaruddin et al., 2013). Additionally, many enzymes, including Gpd1, Dak1, Had1, Gut1, Gpp1, and Gfd1, are involved in glycerol metabolism. Gpd1, a cytosolic form of glycerol‐3‐phosphate dehydrogenase, is important for glycerol accumulation in the appressorium (Foster et al., 2017). Disruption of GPD1 or GPD2 leads to reduced fungal virulence on rice (Shi et al., 2018). Metabolites related to glycerol metabolism including phospholipids could be important for fungal infection.

Phospholipids, as amphipathic molecules, are major components of the cell membrane and play a critical role in various biological processes (Xie, 2019). In M. oryzae, sphingolipid is a type of phospholipid is required for appressorium formation. MoLag1 is involved in the biosynthesis of the sphingolipid of ceramide, and deletion of MoLAG1 causes a severe reduction in conidiation and virulence. This shows that ceramide is important for fungal pathogenicity (Liu, Yun, et al., 2019a; Liu, Liang, et al., 2019b). MoLPPs also encode a type of phosphatidate phosphatases (PAPs), and Molpp3 and Molpp5 mutants display reductions in appressorium formation and virulence that can be restored by exogenous supplementation of diacylglycerol (Skamnioti & Gurr, 2007). Because of the critical role of phospholipids in M. oryzae, key enzymes involved in phospholipid biosynthesis are used as targets of rice blast fungicides. The widely used fungicides phosphorothiolates and isoprothiolane inhibit the transmethylation step of choline biosynthesis (Uesugi, 2001). The impact of phospholipid biosynthesis on virulence has also been reported in the human pathogen Candida albicans, and phosphatidylserine (PS) and phosphatidylethanolamine (PE) are important for pathogenesis (Mu et al., 2019). Deletion of CHO1 causes the loss of PE biosynthesis from PS and in turn attenuates virulence on mice in C. albicans (Cassilly & Reynolds, 2018). Therefore, investigating the genes involved in phospholipid biosynthesis is necessary for understanding the mechanism of fungal virulence and providing potential targets for novel fungicides.

PAPs catalyse the Mg²⁺‐dependent dephosphorylation of phosphatidic acid (PA) to synthetize diacylglycerol (DAG) (Adeyo et al., 2011). Four PAPs, Dpp1, Lpp1, App1, and Pah1, have been elucidated in Saccharomyces cerevisiae (Chae & Carman, 2013; Han et al., 2008; Pascual & Carman, 2013). Furthermore, only four out of eight PAPs have been characterized in M. oryzae, and disruption of MoLPP3 or MoLPP5 results in reduced virulence on rice (Sadat et al., 2014). In S. cerevisiae, the pah1 mutant grows slowly and is hypersensitive to high temperature (Park et al., 2015). Pah1 is involved in the regulation of autophagy via the Nem1/Spo7–Pah1 complex (Rahman et al., 2018). The phosphorylation of Pah1 is regulated by multiple protein kinases in S. cerevisiae (Karanasios et al., 2010).

In this study, we characterized MoPAH1 in M. oryzae. A MoPAH1 deletion mutant was severely defective in vegetative growth, ALS formation, and pathogenicity. RNA sequencing (RNA‐Seq) analysis supported this impact of MoPAH1 deletion on fungal development and pathogenicity. The mutant was hypersensitive to the cell wall stressor Congo red and heat stress, and lipid profiles were widely altered. The PA biosensor indicated that PA is distributed in the hyphae and showed that PA accumulated in the vacuoles of the mutant. Additionally, Pmk1 activation and autophagy were attenuated in the Mopah1 mutant. Overall, the evidence indicates that MoPah1 plays pivotal roles in hyphal growth, conidiation, ALS formation, and pathogenesis in M. oryzae.

2. RESULTS

2.1. Characterization and analysis of MoPah1 in  M. oryzae

To characterize the MoPah1 in M. oryzae, the yeast Pah1 protein sequence was used to search the M. oryzae genome database (Howe et al., 2020) and only one homolog was obtained, named MoPah1 (MGG_01311). MoPah1 contains two protein domains, LNS2 and Lipin_N (Figure S1) (Irie et al., 1993), whereas Pah1 homologs from rice and Arabidopsis contain an extra protein domain, Lipin_mid (Han et al., 2012). Phylogenetic analysis shows that Pah1 is highly conserved, and MoPah1 has a close relationship with Pah1 homologs from filamentous fungi Neurospora crassa and Fusarium graminearum. The MoPah1 protein shows a 70% identity with those of N. crassa and F. graminearum.

High‐quality transcriptomes of M. oryzae during the infection process (Jeon et al., 2020) indicated that expression of PAP genes increases in infected rice tissues, with MoPAH1 the highest (Figure S2), suggesting that MoPAH1 is involved in plant infection.

2.2. The Mopah1 mutant is defective in hyphal growth and conidiation

To investigate the role of the MoPAH1 gene in M. oryzae, we deleted MoPAH1 (Figure S3a). A screen of 120 transformants by PCR identified two Mopah1 mutants, M21 and M22. The Mopah1 mutant M21 was used for further analyses (Table 1 and Figure S3b).

TABLE 1.

Wild‐type and mutant Magnaporthe oryzae strains used in this study

Strain	Genotype description	Reference
WT	Wild‐type strain (P131)	Xue et al. (2012)
M21	MoPah1 mutant of WT	This study
M22	MoPah1 mutant of WT	This study
CP1	MoPah1 complemented transformant	This study
OPW3	Transformant of WT expressing ScOpi1‐RFP	Han and Carman (2017)
OPM5	Transformant of M21 expressing ScOpi1‐RFP	Han and Carman (2017)
EP3	Transformant of WT expressing MoPah1‐GFP	This study
ATG8‐WT5	Transformant of WT expressing MoAtg8‐GFP	This study
ATG8‐M1	Transformant of M21 expressing MoAtg8‐GFP	This study

Strains were grown on complete medium (CM), oatmeal medium (OMA), and minimal medium (MM) plates. The growth rate of the Mopah1 mutant was reduced up to 83% on OMA compared to the wild type (WT) and complementation strains (Figure 1a,b). A cross‐section of colonies on CM plates revealed that mycelia of the Mopah1 mutant grew into the solid medium under the inoculation plug (Figure 1g). Measurements of Calcofluor white (CFW)‐stained hyphae indicated that the average cell length of apical hyphal cells of the Mopah1 mutant was 33% shorter compared to the WT and complementation strains (Figure 1d,e). The aerial hyphae of the Mopah1 mutant also appeared compact compared to the WT (Figure 1h). These results indicate that MoPah1 is involved in vegetative growth.

FIGURE 1.

MoPah1 is involved in vegetative growth and conidiation. (a) Five‐day‐old cultures on complete medium (CM), oatmeal medium (OMA), and minimal medium (MM) plates. The wild‐type (WT), Mopah1 mutant (M21), and Mopah1 complementation transformant (CP1) cultured at 28°C for 5 days (a–c). (b) The colony diameter was measured on CM, OMA, and MM plates. Means and standard deviations were calculated with data from three replicates. (c) Conidiation on OMA plates. Means and standard deviations were calculated from four independent repeats. (d) Hyphal cell length was measured for at least 20 cells for each strain. (e) The septa stained with Calcofluor white (CFW). White arrows indicate cell septa. Bar, 20 µm. (f) The conidiophores were examined under light microscopy. Bar, 50 μm. White arrows indicate the conidiophores. (g) Cross‐section of colonies grown on CM plates for 7 days. (h) The branching of vegetative hyphae in WT and the Mopah1 mutant. Bar, 100 μm. (i) Relative expression levels of conidiation‐related genes in the WT and M21 strains. The expression level of each gene in the WT was set to 1. Means and standard deviations were calculated from three independent replicates. Data were analysed with Student's t test. Asterisks represent significant differences (**p < 0.01, ***p < 0.001, ****p < 0.0001)

Microscopic observations showed that conidiophores of the Mopah1 mutant failed to produce conidia (Figure 1c,f). Conidiation in the complemented strain was similar to the WT. An analysis of the expression of conidiation‐related genes COM1, CON1, CON2, CON6, CON7, COS1, and HTF1 indicated a significant reduction in mutant M21 compared to WT (Figure 1i) (Liu et al., 2010; Shi et al., 1998; Shi & Leung, 1995; Yang et al., 2010). These results indicated that MoPah1 is required for conidiation in M. oryzae.

2.3. Deletion of MoPAH1 alters the lipid profile

MoPAH1 encodes a putative PAP. To assay the function of MoPah1 in vivo, we examined the lipid profile of the Mopah1 mutant by two‐dimensional thin layer chromatography (2D‐TLC) and high‐performance liquid chromatography/mass spectrometry (HPLC‐MS). The 2D‐TLC analysis showed that the lipid profile was altered in the Mopah1 mutant compared with the WT strain (Figure 2a). The lipidomics results using HPLC‐MS showed that the PA level in the Mopah1 mutant was two‐fold higher than in the WT strain. In addition, phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylinositol (PI), and phosphatidylserine (PS) were increased 234%, 93%, 18%, 71%, and 150%, respectively, in the Mopah1 mutant. In contrast, the DAG and triacylglycerol (TAG) content were reduced by 64% and 57%, respectively, in the Mopah1 mutant in comparison to the WT strain (Figure 2b). These results demonstrate that deletion of MoPAH1 widely alters the lipid metabolism in M. oryzae.

FIGURE 2.

Disruption of MoPAH1 widely alters the lipid profile in Magnaporthe oryzae. (a) Two‐dimensional thin layer chromatography (TLC) analysis of total lipids. Total lipids were isolated from the wild‐type (WT) strain and the Mopah1 mutant (M21). Equal amounts of lipids of each sample were loaded on the TLC plate. CL, cardiolipin; DAG, diacylglycerol; PA, phosphatidic acid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PI, phosphatidylinositol; PS, phosphatidylserine; TAG, triacylglycerol. (b) High‐performance liquid chromatography/mass spectrometry (HPLC‐MS) analysis of each lipid in the WT strain and the M21 mutant. Means and standard deviations were calculated from five independent replicates. Data were analysed with Student's t test. Asterisks represent significant differences (**p < 0.01, ***p < 0.001, ****p < 0.0001)

2.4. Disruption of MoPAH1 alters PA distribution

The yeast Opi1 protein (ScOpi1) binds to PA, and its binding domain has been extensively used as a marker for cellular PA (Ganesan et al., 2015). To study the subcellular PA, we fused the PA‐binding domain of ScOpi1 with red fluorescent protein (RFP) and transformed the construct into the WT and Mopah1 mutant strains. The ScOpi1‐RFP protein was detected by western blotting in both the transformed WT (OPW3) and Mopah1 mutant (OPM5) strains (Figure 3b). The protein was not detected in the nontransformed WT control. Under confocal microscopy, red fluorescence was observed in appressoria and invasive hyphae of OPW3 (Figure 3a). In hyphae, fluorescent signals were distributed in the cytoplasm and vacuole (Figure 3c). These results indicate that PA is required for the infection process and accumulated in the vacuole of OPM5. These results suggest that deletion of MoPAH1 affects PA distribution in hyphae, which probably results from the accumulation of PA in the Mopah1 mutant.

FIGURE 3.

ScOpi1‐RFP fluorescence signals in the Mopah1 mutant accumulated in the vacuole. (a) Subcellular localization of ScOpi1‐RFP in appressoria (AP) and invasive hyphae (IH) of the wild type (WT; OPW3). Bar, 10 μm. (b) The ScOpi1 protein level was assessed using western blot analysis in the OPW3 and OPM5 strains. Proteins from the WT strain were used as the negative control. CBB, Coomassie Brilliant blue staining. (c) Subcellular localization of ScOpi1‐RFP in hyphae of the WT (OPW3) and Mopah1 mutant (OPM5) strains. Bar, 10 μm

2.5. MoPah1 regulates responses to multiple stressors

To investigate whether MoPAH1 is involved in stress responses, WT, Mopah1 mutant (M21), and complementation (CP1) strains were grown on CM and treated with stress conditions. These results showed that growth of the Mopah1 mutant was completely inhibited by the high temperature of 34°C. The cell wall‐disturbing reagent Congo red (CR, 0.2  mg/ml) inhibited growth by 33% compared to the WT and CP1 strains. With osmotic stressors NaCl (0.7  M) and sorbitol (1  M), the Mopah1 mutant grew slightly faster than on the CM medium control (Figure 4a,b). In contrast, growth of the WT and complementation strains was reduced by these two stressors. These results indicate that MoPAH1 plays different roles in the response to various stress conditions.

FIGURE 4.

Pleiotropic defects of the Mopah1 mutant in response to stressors. (a) Colony morphology of the wild type (WT), Mopah1 mutant (M21), and Mopah1 complementation (CP1) strains supplemented with different stressors on CM plates. CM, complete medium; CR, Congo red. (b) The inhibition rate was calculated relative to the growth rate of each untreated control. The inhibition extent (%) = [(diameter of untreated colony − diameter of treated colony)/diameter of untreated colony] × 100. Means and standard deviations were calculated from three independent replicates. Data were analysed with Student's t test. Asterisks represent significant differences (****p < 0.0001)

2.6. MoPah1 is required for ALS formation and pathogenicity

ALSs formed at hyphal tips are critical in plant infection for M. oryzae (Kong et al., 2013). Because the Mopah1 mutant does not produce conidia, fresh mycelial blocks were used to test the ability to produce ALSs. When placed on hydrophobic cover slides and barley leaves, the Mopah1 mutant failed to form any ALS after 48  h of incubation (Figure 5a,b). In contrast, the WT and complementation strains formed abundant melanized, dome‐shaped ALSs. Therefore, MoPAH1 is required for the formation of ALSs.

FIGURE 5.

MoPah1 is required for formation of appressorium‐like structures (ALSs) and pathogenicity. (a, b) ALSs formed from hyphal tips on glass coverslips and barley leaves of the wild type (WT) and Mopah1 complementation (CP1) strains, but not the Mopah1 mutant (M21). Bar, 50 μm. (c) Inoculation of rice leaves with mycelial plugs for 14 days. CK, inoculation with agar blocks. (d) Infection assays on detached barley leaves with mycelial plugs for 4 days. CK, inoculation with agar blocks. (e) Statistical analysis of lesion area on rice and barley leaves. Means and standard deviations were calculated from three and five independent replicates in rice and barley leaves, respectively. Data were analysed with Student's t test and asterisks represent significant differences (***p < 0.001, ****p < 0.0001)

To investigate whether deletion of MoPAH1 affects the plant infection capacity of the fungus, we tested the virulence of the WT, M21, and CP1 strains on 4‐week‐old rice seedlings of cultivar Lijiangxintuanheigu and 6‐day‐old barley seedings of cultivar E9. At 14  days postinoculation (dpi) with WT and CP1 strains, plants developed typical large blast lesions on rice leaves. In contrast, M21 caused no visible lesions. On barley leaves, the Mopah1 mutant was completely nonpathogenic and did not produce lesions at 4  dpi, whereas WT and CP1 strains formed typical lesions (Figure 5c). These results indicate that MoPah1 is required for plant infection in M. oryzae.

2.7. Distribution of MoPah1 during developmental and infection processes

The subcellular localization of MoPah1 at different developmental stages was investigated by transforming a MoPah1‐green fluorescent protein (GFP) construct into the WT strain. The expression and integrity of the MoPah1‐GFP fusion protein in transformant EP3 was verified by western blot analysis (Figure S4). Confocal microscopy revealed that the MoPah1‐GFP protein was expressed at all tested developmental stages, including conidia, appressoria, hyphae, and invasive hyphae (Figure S5). The GFP signals appeared mostly in the cytoplasm at different stages. These results suggest that MoPah1 is important during all developmental and plant infection processes for M. oryzae.

2.8. Deletion of MoPAH1 alters the expression of genes involved in development and pathogenesis

To investigate the genome‐wide effect of the MoPAH1 disruption on gene transcriptional expression, we performed RNA‐Seq analysis of the WT and Mopah1 mutant strains. RNA was extracted from 4‐day‐old MM cultures of the WT and M21 strains. RNA samples were sequenced and analysed using an established bioinformatic pipeline (Wang et al., 2009). Approximately 40 million reads were generated from each RNA‐Seq library. Most of the reads mapped (96%) to the reference M. oryzae genome. The expression profile analysis identified 2459 differentially expressed genes (DEGs) with 1035 up‐regulated and 1424 down‐regulated genes from the Mopah1 mutant in comparison with the WT strain (p  ≤  0.05, log₂ FC  ≥  1; Figure 7a). The analysis of the KEGG pathway and GO enrichment showed that autophagy and the metabolic pathways of lipid, glycerolipid, and glycolysis/gluconeogenesis were altered in the Mopah1 mutant (Figures S6 and S7). Based on the functionally characterized genes in the PHI database (Urban et al., 2020), 104 DEGs were categorized into groups representing growth, conidiation, appressorium, and pathogenicity functions (Figure 6b). Genes involved in conidiation and pathogenicity account for the most in the Venn diagram. Analysis by reverse transcription quantitative PCR (RT‐qPCR) confirmed the reduced expression in the Mopah1 mutant of DEGs involved in fungal growth (FLB3, HOX1, HOX4, Dac, CRZ1, RGA4, COD1, CRF1, MSN2), conidiation (TDG1, TDG7, FAR2, CYP51B, CRZ1, RGA4, CHS1, COD1, CRF1, MSN2), appressorium formation (CBP1, CHS7, PTH11, RGA4, CHS1, COD1, CRF1, MSN2), and pathogenicity (RSY1, Dac, CYP51B, PTH11, CHS7, CRZ1, CHS1, COD1, CRF1, MSN2) (Figure 6c). Similar analysis of the strains grown of CM indicated a significant reduction of the expression of conidiation genes in the Mopah1 mutant (Figure 1i). Based on these results, we conclude that MoPah1 is associated with proper expression of genes involved in fungal growth, conidiation, appressorium formation, and pathogenicity in M. oryzae.

FIGURE 7.

Defects of the MAPK signalling pathways in the Mopah1 mutant. (a) Assays of the protein level of Pmk1 and Mps1, phosphorylated Pmk1, and Mps1 were detected using anti‐Phospho‐p44/42 MAPK antibody. Mps1 antibody, Pmk1 antibody and anti‐actin antibody. (b) Analysis of the Pmk1 and Mps1 protein level, phosphorylated Pmk1, and Mps1 relative to that of actin. (c) Relative expression of phosphoproteins against total proteins. The protein contents of phospho‐Mps1, Mps1, phospho‐Pmk1, and Pmk1 relative to that of actin

FIGURE 6.

MoPah1 regulates genes involved in development and pathogenesis in Magnaporthe oryzae. (a) Differentially expressed genes (DEGs) are shown in the volcano plot. The x axis shows the fold change in gene expression between wild type (WT) and Mopah1 mutant M21, and the y axis shows the statistical significance of the difference. Red dots represent up‐regulated genes; blue dots represent down‐regulated genes. (b) Venn diagram of DEGs classified by biological function. The coloured circles represent genes involved in fungal growth, conidiation, appressorium formation, and pathogenicity. The number at the overlapping area stands for DEGs with two or more functions. (c) Reverse transcription quantitative PCR assays of DEGs between the WT and the Mopah1 mutant. The column colour of the WT strain indicates the biological functions shown in (b). Means and standard deviations were calculated from three independent replicates. Data were analysed with Student's t test. Asterisks represent significant differences (***p < 0.001, ****p < 0.0001)

2.9. Pmk1 phosphorylation was reduced in the Mopah1 mutant

The MAPK pathway is required for appressorium formation in M. oryzae (Li et al., 2012). To determine whether the deletion of MoPAH1 altered the MAPK signalling pathway in the Mopah1 mutant, assays for Pmk1 and Mps1 MAP kinase were conducted using western blotting. The protein level of Pmk1, but not Mps1, was reduced dramatically in the Mopah1 mutant. The Pmk1 phosphorylation level was reduced but the Mps1 phosphorylation level increased in the Mopah1 mutant compared to the WT strain. In contrast, the ratio of phosphorylated Pmk1 to Pmk1 was much higher in the Mopah1 mutant than that in the WT, while the same ratio for Mps1 was slightly changed in Mopah1 (Figure 7a,b). Taken together, these results indicate that MoPah1 is mainly involved in the Pmk1 MAPK signalling pathway in M. oryzae, in comparison to the Mps1 MAPK pathway.

2.10. Autophagy is delayed in the Mopah1 mutant

To determine whether MoPah1 is involved in autophagy, MoAtg8, a widely used autophagy marker (Klionsky, 2011), was fused to GFP and transformed into the WT and Mopah1 mutant strains to yield the transformant strains ATG8‐WT5 and ATG8‐M5, respectively. Growth on nitrogen starvation medium‐induced autophagy in both the WT and mutant strains was measured and observed by the increase of free GFP. Surprisingly, autophagy was strikingly delayed in the Mopah1 mutant compared to the WT strain. After 5  h, the fluorescent signals were translocated from the cytoplasm into the vacuole completely in the WT strain (Figure 8a). In contrast, only a portion of fluorescent signals were translocated into the vacuole in the Mopah1 mutant. In immunoblotting assays, ratios of free GFP to the total amount of GFP‐MoAtg8 and free GFP were 90% in the WT strain and only 60% in the Mopah1 mutant (Figure 8b). These results demonstrate that MoPAH1 is required for proper nonspecific autophagy in M. oryzae under conditions of nitrogen starvation.

FIGURE 8.

Autophagy is delayed in mutant Mopah1. (a) The fluorescence signal of GFP‐MoATG8 in the wild‐type strain (WT) and Mopah1 mutant (M21) when cultured in complete medium (CM) or nitrogen starvation medium (MM−N). Bar, 10 μm. (b) Total proteins were extracted from ATG8‐WT5 and ATG8‐M5 strains cultured in MM−N for 0, 3, and 5 h. The protein level of GFP‐Atg8 was detected using immunoblotting. Coomassie Brilliant blue (CBB) staining indicates the protein loading for each lane. (c) Autophagy was delayed in the Mopah1 mutant. The extent of autophagy was analysed by calculating the amount of free GFP compared with the total amount of GFP‐MoAtg8 and free GFP. Quantitative analysis was conducted using ImageJ. Means and standard deviations were calculated from three independent replicates. Data were analysed with Student's t test. Asterisks represent significant differences (**p < 0.01)

3. DISCUSSION

Phospholipids are important for various biological processes but direct evidence of a role for phospholipids in M. oryzae pathogenesis was lacking. Recent studies have shown the versatile roles of PA, the simplest phospholipid molecule, in mammalian cells and plants (Dubots et al., 2012; Kim & Wang, 2020). In this study, we set out to evaluate the role of PA in fungal pathogenesis by focusing on the major PAP, MoPah1, an enzyme for PA metabolism. Our results showed the critical role of MoPAH1 in lipid metabolism, fungal development, and pathogenicity.

Among all PAP knockout mutants, only the Mopah1 mutant showed obvious and pleiotropic phenotypes, suggesting an indispensable role of MoPah1 in M. oryzae (Sadat et al., 2014). Deletion of MoPAH1 resulted in an increased level of PA, which has also been shown in the pah1 mutant of S. cerevisiae (Han et al., 2006). Decreased levels of DAG and TAG were observed in the Mopah1 mutant, which has also been observed in pah1 mutants of S. cerevisiae, F. graminearum, and Yarrowia lipolytica (Karanasios et al., 2010; Oh et al., 2017; Ukey et al., 2020). The reduction of TAG levels leads to a reduced level of lipid droplets in pah1 mutants of Tetrahymena thermophila and F. graminearum (Liu, Yun, et al., 2019a; Liu, Liang, et al.,2019b; Pillai et al., 2017; Shi et al., 2018). The KEGG pathway and GO term analyses of the Mopah1 mutant showed that many DEGs are related to lipid and glycolysis/gluconeogenesis metabolism. These results indicate a conserved function for Pah1 in lipid metabolism in fungi.

The Mopah1 mutant was severely defective in mycelial growth, hypersensitive to cell wall stressors, and completely inhibited by elevated temperatures. These phenotypes are similar to the pah1 mutants of S. cerevisiae, which is hypersensitive to oxidative stress and elevated temperature (Carman & Han, 2019), and to the Fgpah1 mutant of F. graminearum, which is defective in mycelial growth and response to cell wall stressors (Liu, Yun, et al., 2019a; Liu, Liang, et al., 2019b). Furthermore, the down‐regulation of the growth‐related genes MoMSN2 and MoFLB3 suggests their involvement in reduced growth phenotypes of the Mopah1 mutant (Figure 6). As a pathogen adapted to paddy field conditions, heat tolerance is especially important for M. oryzae. We found that HSP60 and HSP10, related to heat tolerance, are up‐regulated in the thermosensitive Mopah1 mutant, indicating that the up‐regulation of these HSP genes is insufficient to compensate the phenotype or dysregulated genes, rather than other commonly known HSPs that may be involved in reduced heat tolerance of the Mopah1 mutant (Li et al., 2011). Conidia are important for plant infection and rice blast spread in the field (Matheis et al., 2017). Unlike the Fgpah1 mutant that produces conidia, the Mopah1 mutant is completely defective in conidiation (Kong et al., 2012; Liu, Yun, et al., 2019a; Liu, Liang, et al., 2019b; Xue et al., 2012). The largest category of DEGs was related to conidiation in the Mopah1 mutant, including transcription factors COM1, COS1, and CON7, confirming the critical role of MoPah1 in conidiation (Figures 1 and 8). Mutants lacking COM1, COS1, or CON7 are also reduced in virulence (Shi et al., 1998; Yang et al., 2010). Taken together, down‐regulated genes involved in development and stress responses in Mopah1 could be also important for fungal virulence.

The Mopah1 mutant is nonpathogenic on plant leaves, similar to the corresponding mutants in C. albicans (Capah1) and F. graminearum (Fgpah1), which are reduced in virulence or are completely avirulent, respectively (Liu, Yun, et al., 2019a; Liu, Liang, et al., 2019b; Mu et al., 2019). The nonpathogenicity of the Mopah1 mutant is probably caused by its defects in ALS formation and the dysregulation of the Pmk1 MAPK pathway, which is responsible for ALS formation (Kong et al., 2013). The ratio of phosphorylated Pmk1 to total Pmk1 was increased in the Mopah1 mutant, though the level of phosphorylated Pmk1 was reduced (Figure 7). Similarly, an increased phosphorylated MAPK relative to corresponding total proteins has been observed in the Lpin2 ^−/− (pah1) mutant of human macrophages (Lordén et al., 2017). In mice, an increased level of PA activates the MEK–Erk MAPK pathway (Nadra et al., 2008). The two‐fold increase of PA content in the Mopah1 mutant could induce hyperphosphorylation of Pmk1, given the dramatically reduced protein level of Pmk1. Delayed autophagy could result in delayed glycogen metabolism, which is possibly another reason for loss of virulence in the Mopah1 mutant (Figure 5). The delayed autophagy could be due to the down‐regulation of the genes MoATG1 and MoATG9 in the Mopah1 mutant (Lv et al., 2017). Additionally, PA is an important component of the cellular membrane system including the vacuole. The S. cerevisiae pah1 mutant lacks vacuolar fusion (Sherr et al., 2017), which plays a critical role in autophagy. Thus, delayed autophagy in the Mopah1 mutant might result from the altered vacuolar membrane due to membrane phospholipid composition changes. Taken together, the altered lipid metabolism in the Mopah1 mutant probably affects Pmk1 phosphorylation activation and autophagy, which subsequently results in complete loss of virulence.

In conclusion, this study has demonstrated that MoPah1, the major PAP, is involved in several developmental and infection‐related processes in M. oryzae. In particular, the MoPah1 PAP is essential for conidiation and ALS formation, indicating that MoPah1 plays a broader role in M. oryzae development than in C. albicans and F. graminearum (Liu, Yun, et al., 2019a; Liu, Liang, et al., 2019b; Mu et al., 2019). MoPAH1 is involved in protein expression and phosphorylation of Pmk1, the most critical MAP kinase in M. oryzae, which advances our understanding of the role of MoPah1 and PA in M. oryzae pathogenicity. However, the detailed mechanism on how MoPAH1 functions is still unknown. The increased level of plant PA negatively modulates relevant transcription factors in their binding to downstream genes (Kim et al., 2019), and similar mechanisms might apply to the Mopah1 mutant, which shows increased PA content and many down‐regulated infection‐related genes (Figures 2 and 8). Notably, down‐regulated genes in the Mopah1 mutant include MoMSN2, COS1, HTF1/HOX2 and CHS7, and 22 of them overlap 62 down‐regulated genes in the mutant lacking MoMSN2, which regulates lipid metabolism, appressorium formation, and invasive growth (Zhang et al., 2014), indicating that down‐regulated MoMSN2 could be related to many phenotypes of the Mopah1 mutant. Additional studies are needed to investigate whether MoPah1 phosphorylation/dephosphorylation regulates the enzymatic activity of the versatile protein, like reports in yeast and other filamentous fungi (Carman & Han, 2019). These studies will provide a clear role of MoPah1 in M. oryzae pathogenicity, particularly on Pmk1 phosphorylation and regulation of important infection‐related transcription factors.

4. EXPERIMENTAL PROCEDURES

4.1. Strains and culture conditions

M. oryzae P131 was used as the WT in this study (Xue et al., 2012). The WT strain and transformants were routinely cultured on OMA or CM agar plates at 28°C. The growth rate was measured on CM, MM, and OMA at 28°C for 5  days (Li et al., 2017; Zhong et al., 2016). Conidiation was assayed on OMA plates and conidiophores were observed as previously described (Dubey et al., 2019). For tolerance tests, fungal strains were cultured at 28°C on CM plates supplemented with 0.2  mg/ml CR, 0.7  M NaCl, or 1  M sorbitol. For heat tolerance assays, the temperature was set at 34 ℃. For transformation, all transformants were selected on TB3 medium with 250  μg/ml hygromycin B (Genview) or 400  μg/ml geneticin (GLPBIO).

4.2. Targeted gene deletion and complementation of MoPAH1

A split‐marker strategy was used to delete the MoPAH1 gene (Catlett et al., 2003). The 1.0  kb upstream and downstream flanking fragments were amplified using primer pairs 1F/2R and 5F/6R, respectively. Primers HYG‐F/R were used to amplify the hph cassette from plasmid pCB1003 (Carroll et al., 1994). The hph cassette was ligated with flanking fragments using primers 1F/HYG‐UR and HYG‐DF/6R. PCR products were directly used to transform into protoplasts of P131 using the polyethylene glycol (PEG)‐mediated approach as described (Yang et al., 2010). Putative deletion transformants were verified by PCR.

For complementation assays, we used primers MoPAH1‐CF/CR to amplify the full length of the MoPAH1 gene, including its 1.5‐kb native promoter, and clone into the pKNTG plasmid (Chen et al., 2014). The constructs were sequenced and transformed into protoplasts of M21. Putative complementation transformants were verified by PCR (Table  S1).

4.3. Phylogenetic analysis

We downloaded Pah1 protein sequences (M. oryzae, C. albicans, S. cerevisiae, Oryza sativa, Arabidopsis thaliana, Schizosaccharomyces pombe, N. crassa, F. graminearum, Aspergillus nidulans, Sclerotinia sclerotiorum, and Botrytis cinerea) from the NCBI database to construct a multispecies neighbour‐joining phylogenetic tree. Multiple amino acid sequences of Pah1 proteins from different species were aligned using ClustalW. The neighbour‐joining phylogenetic tree was constructed using MEGA 7.0 in the p‐distance model with 1000 bootstrap replicates.

4.4. Calcofluor white staining assay

For CFW staining, mycelia from liquid CM culture with shaking (120  rpm) at 28°C for 24  h were harvested and stained with 10  mg/ml CFW (Sigma‐Aldrich) in the dark for 5  min (Guo et al., 2017). Stained mycelia were observed by microscopy (Ni90; Nikon). The cell length was measured using ImageJ software.

4.5. RNA‐Seq analysis

The WT and Mopah1 mutant (M21) strains were grown in liquid MM with shaking (120  rpm) at 28°C for 4  days. Total RNA was extracted using the RNAprep Pure Plant Kit (Tiangen). RNA quantity and quality were assayed on a NanoDrop spectrophotometer ND‐1000 (Thermo Scientific) and 2100 Bioanalyzer (Agilent). Samples were sequenced on a BGISEQ‐500 sequencer (BGI) following the manufacturer's instructions. Quality control of sequencing data was performed using FastQC (Brown et al., 2017). We used HISAT2 v. 2.0.4 to map the clean reads to the M. oryzae MG8 genome (http://fungi.ensembl.org/index.html) (Kim  et al., 2015). Quantification of gene expression was performed using featureCounts (Liao et al., 2014). DESeq2 was used for differential expression analysis between WT and M21 (Love et al., 2014). DEGs were identified with criteria p  ≤  0.05 and log₂FC  ≥  1 (FC, fold‐change) (Oh et al., 2017). The volcano plot was produced from R analysis (Zhou et al., 2021). The genes used in the Venn diagram were retrieved from PHI‐base (Urban et al., 2020) and drawn using Evenn (Chen et al., 2021).

For heatmap analysis, the transcriptional expression levels of PAP‐encoding genes in M. oryzae were obtained from the previous study (Jeon et al., 2020). The heatmap was drawn using TBtools, as previously described (Chen et al., 2020).

4.6. RT‐qPCR

Total RNA was isolated from mycelia grown on CM plates with cellophane overlays and used for both RT‐qPCR assays and RNA‐Seq analysis. For RT‐qPCR assays, 1st Strand cDNA Synthesis Kit (Vazyme) was used to generate complementary DNA (cDNA). The qPCR assay was performed on the CFX Connect Real‐time PCR System (Bio‐Rad) using TransStart Tip Green qPCR SuperMix (TransGen Biotech) (Li et al., 2011). The β‐tubulin gene (MGG_00604) was used as the internal control, and the expression level of target genes was calculated using the 2^{−∆∆ C t} method (Livak & Schmittgen, 2001). The RT‐qPCR assay was repeated three times with independent samples. Data were analysed with GraphPad software. The primers used are listed in Table  S1.

4.7. ALS formation and plant infection assay

To observe ALSs at hyphal tips, fresh mycelial blocks (2  ×  2  mm) were incubated on hydrophobic glass coverslips at 28°C for 48  h. Formation of ALSs was examined under a microscope (Ni90; Nikon) (Liu et al., 2010).

Six‐day‐old barley (Hordeum vulgare) seedlings of cultivar E9 and 4‐week‐old rice (O. sativa) seedlings of cultivar Lijiangxintuanheigu were used for infection assays. Fungal culture blocks were inoculated on barley and rice leaves in a humid chamber at 28°C. Symptoms on rice and barley were examined at 14 and 4  dpi, respectively. Plant inoculation, incubation, and lesion examination were performed as previously described (Guo et al., 2019).

4.8. Lipid analysis

Mycelia were harvested from liquid MM cultured at 28°C with shaking (120  rpm) for 4  days and freeze‐dried. The sample was incubated in hot (75°C) isopropanol containing 0.05% (vol/vol) butylated hydroxytoluene (Sigma‐Aldrich) for 15  min. Total lipid was extracted using chloroform/methanol (2:1) with 0.01% butylated hydroxytoluene, and this extraction procedure was repeated five times. Then 1  M KCl and water were added to the sample for centrifugation, and the upper phase was discarded. The solvent was evaporated under an nitrogen gas stream and redissolved in chloroform (5  mg/ml) (Bligh & Dyer, 1959). TLC plates were incubated at 110°C for 90  min and samples were spotted at one corner of the plates (Macherey Nagel). Chloroform/ethanol/ammonium hydroxide (65:25:2, vol/vol/vol) and chloroform/ethanol/acetic acid/water (85:15:10:3, vol/vol/vol/vol) were used for first‐ and second‐dimension separation, respectively. After drying, the plates were exposed to iodine vapour for 90  s in the tank. The lipid contents were measured as described using an Agilent HPLC system coupled with a triple quadrupole/ion trap 4000 QTrap mass spectrometer (Applied Biosystems) (Nakamura et al., 2014).

4.9. Subcellular localization assays

The MoPah1‐GFP and ScOpi1‐RFP constructs were generated using a One Step Cloning Kit (Vazyme) and verified with sequencing. The constructs were transformed into the WT and Mopah1 mutant strains (Yang et al., 2010). The transformant strains were used for observing GFP and RFP fluorescence signals at different developmental stages, including hyphae, conidia, appressoria, and invasive hyphae, under a confocal microscope (TCS SP8; Leica). Confocal images were collected with 40× objective lenses. Excitation and emission spectra of eGFP and mRFP were 498/550  nm and 571/620  nm, respectively. The images were acquired and processed using LAS X software.

4.10. Western blotting

Total proteins were extracted from mycelia cultured in liquid CM for 48  h as previously described (Li et al., 2011). Proteins were separated with 10% SDS‐PAGE and then transferred to polyvinylidene fluoride (PDVF) membranes. The target proteins on the PDVF membrane were detected with anti‐GFP and anti‐RFP antibodies (Abbkine), along with horseradish peroxidase (HRP)‐conjugated secondary antibodies (Abbkine). Signals were visualized with a Western Blotting Detection Kit (Advansta) and the membrane was photographed using the ChemiDoc Touch imaging system (Bio‐Rad). The protein loading was monitored by staining proteins in acrylamide gels with Coomassie Brilliant blue (CBB) for 2 h and destaining the gel overnight before imaging (Wang et al., 2018).

4.11. Autophagy analysis

For autophagy analysis, the construct of MoAtg8 fused with a GFP tag at its N‐terminus was transformed into the WT and Mopah1 mutant (Veneault‐Fourrey et al., 2006). The GFP‐Atg8‐tagged WT and mutant strains were cultured in CM for 2  days. Mycelia were rinsed with water and transferred to nitrogen starvation medium for 2–5  h to induce nonselective autophagy. The fluorescence signals were observed under a confocal microscope (TCS SP8; Leica). Proteins were extracted and analysed by western blotting with anti‐GFP antibodies as described (Li et al., 2017). The amount of free GFP and GFP‐Atg8 was quantified by densitometric analysis with ImageJ (Zhang et al., 2018; Zhong et al., 2016).

4.12. Assays for the nonphosphorylated and phosphorylated MAP kinases

Mycelia were harvested from liquid CM for 48  h and used for protein extraction as described (Li et al., 2011). Expression and phosphorylation of Pmk1 and Mps1 were detected with Phopho‐p44/42 MAP kinase antibody kits, and anti‐Mps1 and anti‐Pmk1 antibodies (Cell Signaling Technology; ABclonal Biotechnology) following the manufacturer's instructions. The amount of protein was quantified by densitometric analysis with ImageJ (Qu et al., 2021).

CONFLICT OF INTEREST

There is no conflict of interest.

Supporting information

FIGURE S1 Phylogenetic analysis and domain architecture of Pah1 homologs. MoPah1 of Magnaporthe oryzae (XP_003714235.1) and its homologs among different organisms, including Candida albicans (XP_722512.1), Saccharomyces cerevisiae (NP_013888.1), Oryza sativa (XP_015639725.1, XP_015617116.1), Arabidopsis thaliana (NP_001118604.1, NP_199101.1), Schizosaccharomyces pombe (NP_594815.1), Neurospora crassa (CAE76557.1), Fusarium graminearum (XP_011316593.1), Aspergillus nidulans (XP_658406.1), Sclerotinia sclerotiorum (SS1G_06175), and Botrytis cinerea (XP_001548614.2). The phylogenetic tree was constructed using MEGA 7.0 software with 1000 bootstrap replicates and the neighbour‐joining method (left panel). Conserved domains in Pah1 orthologs (right panel) were predicted with the pFam database

FIGURE S2 Expression of MoPAH1 is induced during rice infection. The heatmap shows the gene expression level of PAP‐encoding genes at 0, 18, 27, 36, 45, and 72 h postinoculation (hpi). The gene expression level was calculated from RNA‐seq data using (log₂ [FPKM + 1])

FIGURE S3 Generation of the Mopah1 mutant. (a) Strategy of MoPAH1 gene deletion. The upstream and downstream flanking fragments were amplified with primer pairs 1F/2R and 5F/6R, respectively, and ligated with the hph cassette by double‐joint PCR. Thick arrows indicate the MoPAH1 and hph genes. The locations of the primers are indicated with thin arrows. (b) Mopah1 mutants (M21, M22) were verified using specific primer pairs 3F/4R, 1F/7R, and 8F/6R. WT, wild‐type strain P131

FIGURE S4 Immunoblotting of total proteins isolated from mycelia of the wild type (WT) and the EP3 strain expressing MoPah1‐GFP using the anti‐GFP antibody. Proteins from WT were used as the negative control. Coomassie Brilliant blue (CBB) staining indicates the protein loading of each lane

FIGURE S5 Subcellular localization of MoPah1‐GFP in Magnaporthe oryzae. Subcellular localization of MoPah1‐GFP in hyphae (HY), conidia (CO), and appressoria (AP) for 24 h postinoculation (hpi) and invasive hyphae (IH) in barley for 30 hpi. The GFP was fused towards the C‐terminus of the MoPah1 protein. Bar, 10 μm

FIGURE S6 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and GO term analyses of differentially expressed genes (DEGs) show the functions of MoPAH1 according to biological processes. (a) Top 25 significantly enriched KEGG pathways of DEGs. (b) GO term analysis of DEGs between Mopah1 and wild‐type strains

FIGURE S7 The heatmap shows the distribution of expression levels of DEGs (log₁₀[FPKM + 1]) related to lipid metabolism, glycerolipid metabolism, glycolysis/gluconeogenesis metabolism, and autophagy. Vertical columns indicate wild type (WT) and M21. The colour key represents the Z score

TABLE S1 Primers used in this study

Zhao, J. , Sun, P. , Sun, Q. , Li, R. , Qin, Z. , Sha, G. , et al (2022) The MoPah1 phosphatidate phosphatase is involved in lipid metabolism, development, and pathogenesis in Magnaporthe oryzae . Molecular Plant Pathology, 23, 720–732. 10.1111/mpp.13193

DATA AVAILABILITY STATEMENT

The RNA‐Seq data used in this study can be accessed at the National Center for Biotechnology Information (NCBI) at http://www.ncbi.nlm.nih.gov/ with the accession number PRJNA777402.