Mitochondrial β-oxidation regulates organellar integrity and is necessary for conidial germination and invasive growth in Magnaporthe oryzae


For correspondence. Naweed I. Naqvi: E-mail; or Rajesh N. Patkar: E-mail; Tel. (+65) 6872 7493; Fax (+65) 6872 7007.


Fatty acids stored as triglycerides, an important source of cellular energy, are catabolized through β-oxidation pathways predicted to occur both in peroxisomes and mitochondria in filamentous fungi. Here, we characterize the function of Enoyl-CoA hydratase Ech1, a mitochondrial β-oxidation enzyme, in the model phytopathogen Magnaporthe oryzae. Ech1 was found to be essential for conidial germination and viability of older hyphae. Unlike wild-type Magnaporthe, the ech1Δ failed to utilize C14 fatty acid and was partially impeded in growth on C16 and C18 fatty acids. Surprisingly, loss of β-oxidation led to significantly altered mitochondrial morphology and integrity with ech1Δ showing predominantly vesicular/punctate mitochondria in contrast to the fused tubular network in wild-type Magnaporthe. The ech1Δ appressoria were aberrant and displayed reduced melanization. Importantly, we show that the significantly reduced ability of ech1Δ to penetrate the host and establish therein is a direct consequence of enhanced sensitivity of the mutant to oxidative stress, as the defects could be remarkably reversed through exogenous antioxidants. Overall, our comparative analyses reveal that peroxisomal lipid catabolism is essential for appressorial function of host penetration, whereas mitochondrial β-oxidation primarily contributes to conidial viability and maintenance of redox homeostasis during host colonization by Magnaporthe.


Triglycerides, a glycerol backbone with three esterified fatty acids, are the most abundant form of lipids. Lipases act upon triglycerides releasing one fatty acid molecule at a time, giving rise to diacylglycerol, and eventually to glycerol. While glycerol is converted to dihydroxyacetone phosphate, free fatty acids are activated by esterification with coenzyme A, followed by their oxidative catabolism (β-oxidation), utilization in the synthesis of complex lipids or as an attachment to certain proteins (McIlhinney et al., 1985; Daum et al., 2007). Different organisms have adopted distinct β-oxidation strategies to utilize fatty acids (Kunau et al., 1995; Hashimoto, 2000; Wanders et al., 2001; Feron et al., 2005). Although β-oxidation of fatty acids occurs primarily in mitochondria, catabolism through peroxisomal β-oxidation is also well known in mammals (Kunau et al., 1995; Hashimoto, 2000; Wanders et al., 2001). Similar β-oxidation pathways have been found to be functional in lower eukaryotes and prokaryotes. In fungi, owing to extensive studies in yeasts, it was generally accepted that the β-oxidation of fatty acids was exclusive to the peroxisomes (Hiltunen et al., 2003). However, previous studies have demonstrated the presence of a mitochondrial β-oxidation pathway in Sporidiobolus pararoseus (Feron et al., 2005) and in the filamentous fungus Aspergillus nidulans (Maggio-Hall and Keller, 2004; Hynes et al., 2006).

The importance of fatty acid β-oxidation in fungal pathogenesis is strongly indicated by the upregulation of lipid metabolism-related genes during infection or through the isolation of non-pathogenic strains harbouring mutations in the genes encoding essential components of such catabolic pathways. For instance, transcriptional profiling of the human opportunistic pathogen Candida albicans showed upregulation of the genes involved in β-oxidation during the macrophage sequestration stage of its infection cycle (Prigneau et al., 2003). Similarly, the Isocitrate lyase 1 (Icl1) and Malate synthase 1 (Mls1) enzymes of the glyoxylate cycle were also found to be upregulated during pathogen engulfment by macrophages (Lorenz and Fink, 2001). As the icl1Δ exhibits decreased virulence in a mouse model, survival within the macrophage most likely requires a metabolic reprogramming that involves a switch to lipid catabolism.

Magnaporthe oryzae, the rice blast pathogen, is a filamentous ascomycete that shows complex morphological differentiation during asexual and pathogenic life cycle (Dean et al., 2005). During pathogenic development in Magnaporthe, conidial germ tubes elaborate specialized infection structures called appressoria. Magnaporthe shows rapid, bulk cytoplasmic streaming to mobilize lipid droplets into the incipient infection structures, wherein such lipid stores are utilized during maturation (Thines et al., 2000). While long-chain fatty acids released from triglycerides are utilized through peroxisomal β-oxidation pathways to produce acetyl CoA and energy, glycerol may contribute directly to the development of enormous turgor pressure within the appressoria (de Jong et al., 1997). Mature melanized appressoria mechanically breach the host cuticle (appressorial function), and elaborate fungal invasive hyphae within the host tissue (Howard and Valent, 1996; Mendgen et al., 1996). The point of fungal entry is usually associated with activation of a strong defence response including a rapid burst of reactive oxygen species (ROS) in the invaded host cell. In order to be successful in colonizing the host, the fungus has to cope with and counteract such unfavourable plant defence response, in particular the oxidative stress (Huang et al., 2011).

Although the mechanism of ROS toxicity is not fully understood, imbalance in the mitochondrial redox state is believed to be a major cause of cellular damage. Various antioxidant molecules and enzymes, including the glutathione (GSH)-dependent antioxidant system, provide cellular protection against ROS under pathological and toxicological conditions (O'Donovan and Fernandes, 2000). Limitation in the synthesis and transmembrane transport of GSH in the mitochondria makes the optimal functioning of the mitochondrial GSH system and maintenance of the redox state essential to protect against the ROS-mediated injuries (O'Donovan and Fernandes, 2000).

Several studies have addressed the role of the peroxisomal β-oxidation in fungal pathogenesis. In the corn pathogen Ustilago maydis, deletion of MFE2 (FOX2), which encodes a multifunctional enzyme that catalyses the second and third enzymatic reactions of the β-oxidation pathway, leads to a significant reduction in virulence marked with decreased proliferation of fungal hyphae within the host and impaired teliospore development (Klose and Kronstad, 2006). This work suggests that in U. maydis peroxisomal β-oxidation is required for metabolism of host lipids during in planta pathogenic development. Similarly, Mfp1 (Fox2) has been shown to play an important role in fatty acid metabolism and pathogenesis in Magnaporthe (Wang et al., 2007). Likewise, loss of Peroxin 6 (ClaPex6) in Colletotrichum lagenarium led to defect not only in the development but also in the function of the appressoria (Kimura et al., 2001). Recently, Yang et al. have shown that Carnitine-acylcarnitine carrier protein Crc1 is required not only for fatty acid metabolism but also for appressorial physiology and function in Magnaporthe (Yang et al., 2012). Our previous work on Pex6 and the Carnitine acetyltransferase Pth2 underscores the importance of peroxisome-based fatty acid metabolism in Magnaporthe pathogenesis (Bhambra et al., 2006; Ramos-Pamplona and Naqvi, 2006). Loss of either Pex6 or Pth2 leads to impaired growth on olive oil-containing medium, weakened cell wall, and reduced melanization of the appressoria, which consequently fail to penetrate the host cuticle (Ramos-Pamplona and Naqvi, 2006). Similarly, mitochondrial β-oxidation pathway enzyme Ech1, which catalyses the second step in the breakdown of fatty acids, has been reported to be important for fatty acid utilization in A. nidulans (Maggio-Hall and Keller, 2004). However, a role for mitochondrial β-oxidation pathway in fungal pathogenesis has not yet been explored or reported. Here, we analyse the function of Ech1 in pathogenic development and virulence in M. oryzae and show that the β-oxidation machinery plays an important role in mitochondrial integrity and is essential for proper invasive growth of Magnaporthe during initiation of the devastating blast disease in rice and barley. Last, we compare the functional importance and individual contributions of the mitochondrial and peroxisomal β-oxidation pathways and the glyoxylate cycle towards infection-related growth and pathogenesis of the rice blast fungus.


Identification of the mitochondrial β-oxidation pathway enzyme and its subcellular localization in Magnaporthe

In order to understand the function of mitochondrial β-oxidation during fungal pathogenesis, a M. oryzae orthologue of a specific β-oxidation enzyme was sought. The Magnaporthe genome database [MGD, Broad Institute, USA (Dean et al., 2005)] was then searched using the blast algorithm for sequences related to the A. nidulans EchA (ANIA_05916). The blast search in the Magnaporthe genome yielded eight hits with MGG_12868 showing the highest degree of identity (72%) and similarity (83%) to ANIA_05916. Furthermore, Pfam analysis predicted an Enoyl-CoA hydratase domain only in MGG_12868. Thus, based on high degree of similarity to ANIA_05916 and the presence of the ECH domain, MGG_12868, hereafter Ech1, was selected for further characterization (Fig. 1A). Analysis of the Ech1-related sequences from various fungi showed that Magnaporthe Ech1 is most closely related to Neurospora crassa EchA (Fig. S1).

Figure 1.

Subcellular localization and expression analysis of Ech1 in M. oryzae.

A. Domain architecture of Ech1 sequence in Magnaporthe.

B. Epifluorescence microscopy-based analysis of MTS-GFP localization during vegetative and pathogenic development in Magnaporthe. The MTS-GFP strain was co-stained with MitoFluor Red 594 to aid visualization of mitochondria during indicated growth stages. Bar, 5 μm.

C. Real-time qRT-PCR analysis showing relative transcript levels of ECH1 or FOX2 in Magnaporthe grown on fatty acids (rice bran oil) at the indicated time points. The transcript levels are represented as fold induction with respect to those from culture grown in glucose-containing medium. The data represent mean ± SEM of two independent experiments conducted in triplicates.

To determine the subcellular localization, we tagged Ech1 with GFP at the C-terminus (Ech1-GFP) in the wild-type background. However, the ECH1-GFP allele was judged to be a hypomorph and compromised for mitochondrial function and hence not used in subsequent experiments. A consensus mitochondrial targeting sequence (MTS) was predicted in Ech1 based on previous studies (Kanazawa et al., 1993) and an assessment by PSORT II analysis. To determine the specificity of the predicted MTS, Ech11–29-GFP was expressed in wild-type Magnaporthe. Localization of Ech11–29-GFP showed punctate and tubular structures in conidia, germ tubes, appressoria, and vegetative hyphae of wild-type Magnaporthe. Colocalization of Ech11–29-GFP with MitoFluor Red 594-stained punctate and tubular mitochondria confirmed the MTS in Ech1 and its targeting to mitochondria in Magnaporthe (Fig. 1B). Hereafter, Ech11–29-GFP is referred to as MTS-GFP. The localization studies thus indicated that Magnaporthe contains both vesicular as well as tubular mitochondria during vegetative and pathogenic differentiation. Similarly, we identified a Magnaporthe orthologue of peroxisomal Mfe2, hereafter Fox2, and characterized it along with Ech1 to evaluate the contribution of such lipid catabolism pathways in the development and pathogenesis of the blast fungus.

Expression of peroxisomal and mitochondrial β-oxidation genes in response to fatty acids

We performed a comparative analysis of the transcript levels of ECH1 and FOX2 at the early (2 and 12 h post inoculation, hpi) and late (24 hpi) stages of vegetative growth in the presence of glucose or fatty acids (rice bran oil). Transcriptional profiling using real-time qRT-PCR at early stage of vegetative growth on fatty acid medium (2 hpi) showed that both ECH1 and FOX2 transcripts were induced by 1.6 ± 0.27 and 3.11 ± 0.16 fold, respectively, when compared to those in the cultures grown in glucose medium (Fig. 1C). Interestingly, at 12 hpi, transcript levels of ECH1 reduced to 1.13 ± 0.05 fold, while those of FOX2 induced further to 5.96 ± 0.15 fold when compared to those at 2 hpi and against levels in the cultures from glucose medium (Fig. 1C). At 24 hpi, both the ECH1 and FOX2 transcript levels reduced to 0.53 ± 0.08 and 4.38 ± 0.25 fold, respectively, when compared to their levels at early growth stages. These results indicate that the β-oxidation enzymes from mitochondria (Ech1) and peroxisomes (Fox2) are regulated in a temporal manner during utilization of long-chain fatty acids.

ECH1-deletion and characterization of the mutant in Magnaporthe

A gene-deletion mutant was generated by replacing the entire open reading frame (ORF) with the selection marker for Bialaphos resistance (Bar), to study the biological function of Ech1 in Magnaporthe (Fig. S2A). Molecular confirmation of the targeted gene deletion was performed by Southern blot analysis. Hybridization with an ECH1 5′ untranslated region (UTR) as a probe detected a band of 1.5 kb and 7 kb in the wild type and the ech1Δ strain, respectively, indicating a successful gene replacement event (Fig. S2B). Genetic complementation of the ech1Δ strain was performed by integrating a PCR fragment encompassing the entire ECH1 ORF with 1.85 kb of 5′ and 0.15 kb of 3′ UTR sequences. Southern blot analysis confirmed the presence of the diagnostic 1.5 kb fragment in the complemented ech1Δ strain (ECH1+, Fig. S2B). Similarly, a fox2Δ mutant was generated and confirmed by Southern blot analysis (Fig. S3).

We evaluated the overall fitness of the ech1Δ mutant by assessing its vegetative growth. Colony growth of the ech1Δ was comparable to that of the wild type, ECH1+ or fox2Δ mutant when seeded using mycelial plugs (Fig. 2A). However, when initiated with conidia, the ech1Δ colony was much smaller compared to the other three aforementioned strains, thus indicating the importance of mitochondrial lipid catabolism in conidial germination and/or hyphal extension (Fig. 2A). Furthermore, conidial germination was found to be significantly reduced (by 25–30%; Fig. 2B; P < 0.005) in the ech1Δ when compared to the wild type, ECH1+ or fox2Δ mutant in Magnaporthe. Although the diameters of the wild type and ech1Δ colonies were comparable when inoculated with mycelial plugs, there was a visible difference in the vegetative hyphae especially in the centre of the colony. Hence, we assessed hyphal viability in the ech1Δ and wild type, using the vital stain Phloxine B. Indeed, Phloxine B-stained colonies showed comparatively enhanced cell death only in the older hyphae in ech1Δ (Fig. 2C) likely due to mitochondrial dysfunction in the mutant.

Figure 2.

Phenotypic characteristics of the ech1Δ mutant in Magnaporthe.

A. Vegetative growth defect of ech1Δ strain. A conidial suspension or mycelial plug of wild type (1), ech1Δ (2), ECH1+ (3) or fox2Δ (4) was inoculated on to prune agar medium. Colony characteristics were assessed and documented at 5 dpi. Arrowheads highlight the difference in the colony diameters of the ech1Δ mutant.

B. Conidial germination in the ech1Δ strain. Approximately, 1000 conidia per indicated strain were inoculated on hydrophobic glass coverslips and the number of germinated conidia quantified after 5 hpi. The data represent mean ± SEM from three replicates.

C. Viability tests of the ech1Δ strain. Mycelial plug of wild type or ech1Δ was grown on CM alone or CM with 10 μM solution of Phloxine B. The colonies were photographed after 7 days.

D. Utilization of different fatty acids by ech1Δ or fox2Δ. The strains [wild type (1), ech1Δ (2), fox2Δ (3) or ECH1+ (4)] were subcultured on basal medium containing fatty acids of different chain lengths (C14, myristic acid; C16, palmitic acid; or olive oil) individually as sole carbon source. The growth was assessed after 7 dpi.

We tested the ability of ech1Δ to utilize fatty acids as the sole carbon source. Although both ech1Δ and fox2Δ showed reduced growth on olive oil-containing medium, ech1Δ was severely affected on C14 fatty acid when compared to wild type or fox2Δ strain (Fig. 2D), indicating that mitochondrial β-oxidation plays an important role in catabolism of long-chain (C14 to C16) fatty acids in Magnaporthe.

Role of Ech1 during pathogenic development in Magnaporthe

To determine the role of Ech1 during pathogenesis, we evaluated the infection-related development of the ech1Δ strain. The ech1Δ and fox2Δ strains developed a significantly high percentage (29–34%, P < 0.005) of aberrant appressoria when compared to the wild type and ECH1+ (Fig. 3A and B). While 95% wild-type or ECH1+ appressoria were spherical, highly pigmented and contained small vesicles distributed in the cytoplasm, the aberrant appressoria of the ech1Δ mutant were misshapen and less melanized (Fig. 3A). Similarly, the aberrant appressoria of the fox2Δ strain were less pigmented and contained large vesicles (Fig. 3A). To check if energy generation through anaerobic catabolism could bypass requirement for β-oxidation, we tested appressorial development in the presence of glucose. Addition of high concentration (2.5%) of glucose at the start of conidial germination resulted in normal appressoria formation at nearly wild-type frequency in both the mutants (Fig. 3B; P < 0.005). This indicates that Magnaporthe is highly dependent on the β-oxidation machinery for energy and/or important metabolic intermediates, particularly while growing under nutrient-deficient conditions on the host surface.

Figure 3.

Defects in appressorial development in the ech1Δ strain in Magnaporthe.

A. Conidia of wild type, ech1Δ, ECH1+ or fox2Δ strain were inoculated on an inductive surface and assessed for appressorium development after 24 h. Arrowheads and numbers represent aberrant appressoria. Bar, 10 μm.

B. Normal spherical and melanized appressoria (out of total appressoria formed) were quantitatively assessed at 24 hpi in conidia untreated (grey) or treated with 2.5% glucose (black column). The data represent mean ± SEM from three independent assays.

C. Appressorial melanization in the ech1Δ strain. Ultrathin sections were taken from the wild type, ech1Δ or fox2Δ appressoria on rice leaf surface at 24 hpi, and observed using a transmission electron microscopy. Bar, 0.5 μm. Arrowheads depict melanin layer between the cell wall and the cell membrane. Inset, complete picture of the appressorium on host surface. The asterisk shows retarded growth of the ech1Δ penetration peg compared to that in the wild type. HCW, host cell wall.

D. Melanin ghosts of the ech1Δ and fox2Δ appressoria. The appressoria from the wild type, ech1Δ or fox2Δ strain were treated to obtain melanin ghosts (see Experimental procedures) at 24 hpi. Tricycl refers to Tricyclazole. Bar, 10 μm.

We further tested ech1Δ and fox2Δ for sensitivity towards calcofluor white (CFW) that interferes in cell wall synthesis, and by cytorrhysis assays to indirectly estimate the appressorial turgor. None of the aforementioned mutants showed any significant anomaly in appressorial turgor (Fig. S4; P = 0.1) or sensitivity towards CFW (Fig. S5), indicating that mitochondrial catabolism of fatty acids does not contribute significantly to cell wall biogenesis and/or generation of appressorial turgor. Our previous work on Magnaporthe Pex6 and Pth2 strongly suggested a role for peroxisomal β-oxidation in the synthesis of appressorial melanin, which is essential for pathogenesis in the blast fungus (Ramos-Pamplona and Naqvi, 2006). Therefore, we assessed appressorial melanin deposition in the ech1Δ by transmission electron microscopy or using melanin ghosts prepared from the appressoria. In the wild-type appressoria, a distinct, uniform electron-dense layer, which corresponds to the melanin layer, was clearly seen between the cell wall and plasma membrane (white arrow, Fig. 3C), whereas the melanin layer was noticeably thinner in the ech1Δ and fox2Δ appressoria (Fig. 3C). Moreover, both the ech1Δ and fox2Δ appressoria showed accumulation of unutilized lipid bodies, which were largely absent in the wild-type infection structures (Fig. 3C). Interestingly, unlike wild-type, which developed healthy penetration pegs and invaded the host cells by 24 to 28 hpi, fox2Δ appressoria were completely impaired in breaching the host cuticle (insets, Fig. 3C). Importantly, although few ech1Δ appressoria could invade the host, the penetration pegs developed were highly unhealthy (inset, asterisk, Fig. 3C). To prepare melanin ghosts, appressoria were treated with guanidinium isothiocyanate and HCl, which yielded particulate suspension. Light microscopy of the particulate suspension revealed structures resembling appressoria, which were highly melanized in case of wild type (Fig. 3D). In contrast, treatment of the ech1Δ or fox2Δ appressoria yielded light coloured suspension, which showed less pigmented melanin ghosts comparable to those obtained from the wild-type appressoria treated with Tricyclazol, a known inhibitor of melanin biosynthesis (Fig. 3D).

It has been shown that there are interconnections of lipidic pathways involved in the regulation of ergosterol biosynthesis (Veen and Lang, 2005). Therefore, we detected ergosterol distribution in the ech1Δ appressorial membranes by Filipin staining. While wild-type appressoria were stained intensely with Filipin, those of the ech1Δ showed a faint fluorescence signal after Filipin staining indicating reduced ergosterol levels in the plasma membrane of the mutant (Fig. S6). Altogether, these results indicate that mitochondrial β-oxidation is important for proper appressorial development and function in Magnaporthe.

Mitochondrial β-oxidation and pathogenesis in Magnaporthe

Considering the defects in the development of appressoria, we tested the ability of the ech1Δ strain to cause disease on rice and barley. The ech1Δ and fox2Δ strains failed to elicit any visible disease lesions on detached barley leaf assays (Fig. 4A). In contrast, inoculation with the wild-type or ECH1+ conidia resulted in the formation of typical blast disease lesions at 5 days post inoculation (dpi) (Fig. 4A). Furthermore and unlike the wild type, the ech1Δ failed to cause disease on rice leaves at 7 dpi (Fig. 4B). In parallel, the wild type caused root infection, which was evident with visible cell death (browning of the roots) and restricted root length at 10 dpi (Fig. 4C). However, neither necrosis nor cell death was seen at 10 dpi in the rice roots inoculated with the ech1Δ strain (Fig. 4C). To understand the reason behind the inability of the ech1Δ strain to cause disease lesions on leaf tissue, we assessed appressorial function of host penetration, which is marked with callose deposition by the host at every site of entry by the fungus. At 48 hpi, 70–75% of the wild-type or ECH1+ appressoria formed penetration pegs (induction of callose deposition) on host leaves. In contrast, approximately 20% of ech1Δ and 2% of the fox2Δ appressoria were capable of penetrating the rice leaves (Fig. 4D; P < 0.005). Surprisingly, although around 20% of the ech1Δ appressoria invaded the host tissue, the mutant remained significantly non-pathogenic. To investigate if this was due to restricted in planta growth, we tested an ech1Δ strain expressing cytoplasmic GFP (Figs S7 and S8) within rice leaf sheath to aid visualization of the invasive hyphae. Consistent with the previous result, around 20% of the cytoplasmic GFP-expressing ech1Δ appressoria, as against ≥ 60% of those of wild type or ECH1+ strain, were capable of entering the host cell at 30 hpi (upper panels, Figs 4E and S9). Interestingly, at 52 hpi, the ech1Δ penetration pegs showed restricted growth with significantly weak cytoplasmic GFP signal indicating non-viability in the penetration hyphae that were still in the primary infected cells (middle panels, Fig. 4E). In contrast, the wild type or ECH1+ strain expressing the cytoplasmic GFP showed penetration hyphae transiting from one host cell to another at 52 hpi (middle panels, Figs 4E and S9). Such wild-type or ECH1+ invasive hyphae colonized the neighbouring cells by 76 hpi (lower panels, Figs 4E and S9). However, even at 76 hpi, the ech1Δ penetration hyphae did not recover and consistently showed diminished cytoplasmic GFP signal and were restricted within the primary host cell invaded (lower panels, Fig. 4E). These results indicate that although partly successful in entering the host, the ech1Δ strain failed in proper colonization of the host tissue. Therefore, we construe that mitochondrial fatty acid metabolism plays an important role in appressorial function in Magnaporthe, and is also essential for proper and sustained invasive growth within the host tissue.

Figure 4.

Pathogenicity defects in the ech1Δ strain.

Conidial suspension containing approximately 1000 spores per 20 μl (one drop) of the indicated strains was inoculated on to barley (A) or rice (B) leaf blades and observed at 5 to 7 dpi. Arrowheads show points of inoculation on the rice leaf pieces.

C. Rice root infection assay with the wild-type or the ech1Δ conidia. Roots of 21-day-old CO39 seedlings were dipped in 1 ml of suspension containing 104 conidia for 15 min followed by plantation in soil (pre-autoclaved) mixed with the remaining spore suspension. Roots were observed and photographed at 10 dpi. Arrowheads depict diseased area in the inoculated root.

D. Quantification of appressorial function in the wild type, ech1Δ, ECH1+ or fox2Δ strain. Barley leaf blades were inoculated with the conidia of the indicated strains for 36 h and stained with aniline blue to aid visualization of the callose deposited by the host. The number of penetration sites (inducing callose deposition) were counted and expressed as percentage of appressoria successful in host entry (appressorial function). The data represent mean ± SEM from three independent experiments.

E. In planta growth of the ech1Δ strain. Rice leaf sheath was inoculated with the cytoplasmic GFP-expressing wild type or ech1Δ and assessed for in planta growth by epifluorescence microscopy at the indicated time points. Asterisks depict appressoria and arrowheads indicate penetration pegs or invasive hyphae. Bar, 10 μm.

Role of Ech1 in redox homeostasis during pathogenesis in Magnaporthe

It has been demonstrated that as fungal pathogen starts invading, the host initiates a defence response that mainly involves build-up of elevated levels of ROS (Huang et al., 2011). Therefore, to check the ability to withstand such oxidative burst, we initially tested the sensitivity of ech1Δ towards menadione that induces ROS accumulation (Criddle et al., 2006). While the wild type or ECH1+ strain could tolerate up to 50 μM menadione in the growth medium, the ech1Δ showed a complete cessation of growth (Figs 5A and S10). Therefore, we extrapolated our results to investigate whether sensitivity to ROS led to growth retardation of ech1Δ in planta. Considering such enhanced sensitivity towards ROS (menadione), we tested whether supplementation with an antioxidant could support in planta growth of the ech1Δ strain. We added antioxidant (GSH or ascorbic acid) to the ech1Δ appressoria on rice leaf sheath at 22 hpi and assessed invasive growth at 48 hpi. Supplementation with either GSH or ascorbate significantly improved the in planta growth, which was evident from successful invasion and proliferation of the ech1Δ strain within the host tissue (by 30%, P < 0.05, Figs 5B and 6). Recently, a likely peroxisome-associated alanine: glyoxylate aminotransferase (Agt1) was characterized in M. oryzae, and the appressorium formation defect of agt1Δ was found to be completely suppressed by exogenous NAD+ and pyruvate (Bhadauria et al., 2012). Therefore, we tested whether such metabolic intermediates could also suppress host penetration and/or colonization defects in the peroxisomal and mitochondrial β-oxidation mutants. Exogenously supplied NAD+ and pyruvate restored host penetration in about 30% of fox2Δ appressoria (P = 0.005). However, the host colonization defects in the ech1Δ mutant could not be suppressed significantly via addition of NAD+ and pyruvate (Fig. 6). On the other hand, unlike in ech1Δ, exogenous application of GSH did not suppress host penetration defect in the fox2Δ mutant (Fig. 6). Altogether, we conclude that while certain metabolic intermediates (such as acetyl CoA and NAD+/pyruvate) generated via peroxisomal lipid catabolism are necessary for appressorial maturation and host penetration, the mitochondrial β-oxidation machinery plays a key role in neutralizing the ROS encountered at the post-penetration stage and thus helps in successful invasion and host colonization by Magnaporthe.

Figure 5.

Sensitivity of the ech1Δ strain towards oxidative stress.

A. Wild type or the ech1Δ strain was grown on CM or CM with indicated concentrations of menadione for 5 days.

B. In planta growth of the ech1Δ (expressing Cyt-GFP) alone or in the presence of reduced GSH or ascorbate. Rice sheath was inoculated with the ech1Δ (Cyt-GFP) conidia and either GSH or ascorbate added at 22 hpi. The growth of the invasive/penetration hyphae of the mutant was observed at 48 hpi. The asterisks depict appressoria and the arrowheads mark the invasive hyphae. Bar, 10 μm.

C. Localization of Pwl2-RFP to the BIC in the invasive hyphae of Magnaporthe. Rice sheath was inoculated with the wild type or ech1Δ strain in the absence or presence of 5 mM GSH and observed at indicated time. The BICs are indicated with arrowheads and the appressoria marked with asterisks. Bar, 10 μm.

Figure 6.

Effect of NAD+ and pyruvate or GSH on host penetration and/or colonization by the β-oxidation mutants of M. oryzae.

Rice sheath was inoculated with either ech1Δ or fox2Δ conidia and supplemented with 5 mM each of NAD+ and pyruvate, or 5 mM GSH at the indicated time points and assessed for invasive growth after 48 hpi.

A. Inoculated rice sheath was fixed with methanol and stained with acid fuchsin to aid visualization of fungal structures. Arrowheads depict invasive hyphae and asterisk denotes representative appressorium. Bar, 10 μm.

B. Column chart representing quantification of appressoria showing invasive hyphae (% appressorial function). Data represent mean ± SEM from two replicates. Ctrl, control; Pyr, pyruvate.

In Magnaporthe, invasive hyphal growth from the penetration site to the neighbouring host cells requires the secretion of effector molecules via highly specialized structures called the biotrophic interfacial complex (BIC), which is followed by effector translocation into the host cytoplasm, likely to prepare the host cells for subsequent invasion (Khang et al., 2010). This confers a crucial role to the BIC during invasive growth of Magnaporthe. Therefore, we checked if the growth of primary invasive hyphae in ech1Δ was restricted prior to BIC formation, by studying the Pwl2-RFP localization therein. While Pwl2-RFP fusion protein localized to the BIC in the wild-type strain at ∼28 hpi, it was not detected even after 48 hpi in the ech1Δ strain indicating that the mutant was restricted prior to BIC formation (Fig. 5C). Importantly, Pwl2-RFP showed proper BIC localization in the invasive hyphae at ∼30 hpi when exogenous GSH was supplied to the ech1Δ prior to host penetration (Fig. 5C). We conclude that an intact mitochondrial fatty acid metabolism is necessary for proper growth of the primary invasive hyphae prior to BIC formation during establishment of in planta pathogenic growth phase in Magnaporthe.

Mitochondrial β-oxidation and the organellar integrity

We tested whether loss of Ech1 function, in addition to its effect on viability in the older vegetative and in planta invasive hyphae, had any effect on the integrity of mitochondria, as β-oxidation is a key metabolic pathway of the organelle. We assessed mitochondrial morphology by studying Magnaporthe strains expressing α-subunit of F1 sector of mitochondrial F1F0 ATP synthase (Atp1)–GFP fusion protein (Fig. 7A) or by staining with MitoFluor Red 594 dye (Fig. 7B) or by transmission electron microscopy of ultrathin sections of the mycelia of the ech1Δ or wild-type strains (Fig. 8A). The wild-type or ECH1+ mitochondria predominantly appeared as fused and elongated tubular network or punctate structures, which were highly dynamic during pathogenic (Fig. 7A) and vegetative growth (Fig. 7B; Movies S1 and S3). In contrast, the ech1Δ mitochondria were punctate/vesicular and either static or showed highly restricted/localized movements when compared to the wild type (Fig. 7; Movie S2). Mitochondrial morphology and dynamics in the fox2Δ were comparable to those in the wild type or ECH1+ (Fig. 7B; Movie S4). These observations on mitochondrial morphology were supported by the electron micrographs where unlike wild type, ECH1+ or fox2Δ, which contained both tubular and vesicular mitochondria, the ech1Δ displayed only the latter morphology (Fig. 8A). In addition, the cristae of the ech1Δ mitochondria were significantly less in number and aberrant (Fig. 8A). To test if the loss of mitochondrial integrity was also associated with impaired invasive growth, we investigated Atp1–GFP localization during in planta differentiation. Indeed, loss of mitochondrial integrity was seen in the primary invasive hyphae of the ech1Δ (Fig. 8B). In contrast, the wild type demonstrated the typical mitochondrial network in invasive hyphae that successfully colonized the host tissue (Fig. 8B). Thus, these results indicate that the impairment in the appressorial function, and non-viability in the invasive and older vegetative hyphae was most likely due to loss of mitochondrial integrity and function, triggered in turn due to lack of Ech1.

Figure 7.

Mitochondrial morphology in the ech1Δ strain.

A. Wild type or ech1Δ strain expressing Atp1–GFP fusion protein was inoculated on an inductive surface and assessed for mitochondrial morphology by epifluorescence microscopy using GFP optics at 0 h (conidia), 4 hpi (hooking) and 16 hpi (appressoria). Ap, appressorium; Co, conidium; Gt, germ tube. Bar, 10 μm.

B. Vegetative hyphae of the indicated strains were stained with MitoFluor Red 594 and observed under an epifluorescence microscope using RFP optics. Bar, 5 μm.

Figure 8.

Mitochondrial morphology during vegetative and post-penetration growth in Magnaporthe.

A. Ultrathin sections of the wild type, ech1Δ, ECH1+ or fox2Δ mycelia grown in liquid CM were observed using a transmission electron microscopy. M, spherical and tubular mitochondria with normal cristae; m, spherical mitochondria with significantly reduced and swollen cristae. Bars, 0.5 μm.

B. Wild type or ech1Δ strain expressing Atp1–GFP was inoculated onto rice leaf sheath and observed for mitochondrial morphology in the invasive hyphae at the indicated time points. Rice leaf sheath was stained with FM4-64 prior to observation to aid visualization of the host cell membrane. Arrowheads depict invasive hyphae, and the asterisk denotes an appressorium. Bar, 10 μm.

In conclusion, while overall compartmentalized lipid catabolism is essential for Magnaporthe pathogenesis, the peroxisomal β-oxidation contributes metabolic intermediates required for proper appressorial maturation and function, whereas the mitochondrial pathway plays a critical role in conidial germination and in maintaining organellar integrity and redox status during host colonization.


Triglyceride mobilization is induced by nutrient starvation, which is relevant to the pre-penetration stage of pathogenesis in Magnaporthe as conidial germination and appressorium development takes place in the absence of exogenous nutrients. Indeed, lipid droplets are abundant in asexual conidia of Magnaporthe (Weber et al., 2001) and their mobilization from conidia to the appressoria and utilization therein is sufficient to fulfil energy and biosynthetic requirements for appressorial development and function. Here, we studied the importance of one of the mitochondrial β-oxidation enzymes Ech1 during Magnaporthe pathogenesis, and found that it is not only involved in proper appressorial development but also plays a key role during in planta growth and in maintaining the organellar integrity in the fungus. To evaluate contributions of different β-oxidation pathways towards fatty acid utilization and pathogenic development in Magnaporthe, we also studied peroxisomal enzyme Fox2.

The human and rat Ech enzymes are composed of 290 amino acids with a putative N-terminal MTS of 29 residues (Minami-Ishii et al., 1989; Kanazawa et al., 1993). A consensus mitochondrial targeting peptide sequence (1 to 29 aa) was identified at the N-terminus of all the Ech1-related protein sequences from different organisms (Fig. S1). The subcellular localization of the Aspergillus EchA was demonstrated to be in the mitochondria in Aspergillus (Maggio-Hall and Keller, 2004). Similarly, Ech11–29 (MTS) localized to the tubular and punctate mitochondria in Magnaporthe. The glutamic acid residues at positions 144 and 164 are conserved in the members of hydratase/isomerase families and are required for the catalytic activity of Enoyl-CoA hydratases (Muller-Newen et al., 1995; Kiema et al., 1999). Glu144 and Glu164 are conserved in Magnaporthe Ech1 (Fig. S1, asterisks). Likewise, Fox2, a well-characterized peroxisomal multifunctional protein, is conserved in yeasts [Fox2 in Saccharomyces cerevisiae (Hiltunen et al., 1992); Mfe2 in Yarrowia lipolytica (Smith et al., 2000)], filamentous fungi [Fox2 in N. crassa and FoxA in A. nidulans (Thieringer and Kunau, 1991; Maggio-Hall and Keller, 2004)] and mammals [Mfe2 (Jiang et al., 1996)].

Dynamic changes in lipid metabolism were evident from gene expression profiling of Blumeria graminis during pathogenesis (Both et al., 2005). The transcripts of genes involved in fatty acid catabolism were abundant in the pre-penetration stages of B. graminis, but were downregulated in the epiphytic hyphae and infected epidermis samples (post-penetration stages). The most striking reduction, 16-fold, was observed for Ech1 transcript (Both et al., 2005). Similar alteration was seen in the mRNA levels of ECH1 and FOX2, in wild-type Magnaporthe grown in the presence of fatty acids as the sole carbon source. While transcript levels of Magnaporthe FOX2 remained upregulated consistently, those of ECH1 were induced initially and downregulated upon continued growth on fatty acids. Here, the data represent a fold change in the ECH1 transcript level in the vegetative hyphae not just with respect to that of FOX2 but also when compared to growth on glucose-containing media. Therefore, although ECH1 transcript levels appear less abundant compared to FOX2, there was a significant increase in ECH1 mRNA during growth on fatty acids when compared to the levels observed in glucose-replete conditions. These observations indicate that the β-oxidation machineries in mitochondria and peroxisomes likely contribute differentially in a temporal manner in Magnaporthe. However, it remains to be seen whether such differential induction occurs during invasive growth in planta.

It has been shown that short- (C5 and C6) or medium- (C8 to C12) chain fatty acids may serve as sole carbon source to the wild-type A. nidulans (Maggio-Hall and Keller, 2004; Hynes et al., 2008). However, short-chain fatty acids were extremely toxic to wild-type Magnaporthe even in the presence of glucose (data not shown). Thus, our observations suggest some key differences in metabolic preferences between A. nidulans and M. oryzae. In addition, Magnaporthe exhibited only sparse growth on long- (C14, C16, C18 and C22) chain saturated fatty acids. However, Magnaporthe is able to efficiently utilize long-chain unsaturated fatty acids, oleic acid (C18:1Δ9c). This indicates that Magnaporthe may preferentially metabolize unsaturated oversaturated fatty acids, and such fatty acids may serve as better carbon source when comparing growth defects in the mutant strains. Wang et al. have shown that Magnaporthe Mfp1 (Fox2) is required for metabolism of long-chain fatty acids such as olive oil or oleic acid (C18) as sole carbon source (Wang et al., 2007). Surprisingly, the fox2Δ mutant generated in the current study was capable of utilizing olive oil as a sole carbon source. Such disparity in the two studies could be due to: (i) differences in the growth media used (minimal medium versus basal medium or defined complex medium) and (ii) different wild-type backgrounds (Guy11 versus B157) used to create the mutants. Here, severely reduced growth of the ech1Δ on < C18 fatty acids indicates different preferential chain length for these two organellar pathways. Interestingly, while wild-type Magnaporthe failed to grow properly on acetate-containing medium, acetate led to the transcriptional upregulation of ECH1 and FOX2 (Fig. S11). Between ECH1 and FOX2, the transcript levels of the latter were induced several fold by acetate. On the other hand, acetate led to a modest and minor induction of ECHA and FOXA, respectively, in A. nidulans (Maggio-Hall and Keller, 2004). Thus, it is possible that the difference in lifestyle, i.e. pathogenic versus non-pathogenic, of these fungi likely influences the metabolic properties and utilization of fatty acids as carbon source. Importantly, mutation in either the first or second step of the β-oxidation pathway (disruption of short-chain acyl-CoA dehydrogenase, ΔscdA or ΔechA respectively) conferred inability to utilize isoleucine and valine as sole carbon source in A. nidulans (Maggio-Hall et al., 2008). However, it remains to be seen if Magnaporthe Ech1 is required for utilization of such amino acids as sole carbon source.

As a further assessment of phenotype, we compared the conidiation efficiency of these mutant strains to the wild type. Conidiation in the fox2Δ was reduced by 30% when compared to the wild type. In contrast, the ech1Δ mutant exhibited proper conidiation comparable to the wild-type levels (data not shown). Thus, it is most likely that loss of mitochondrial β-oxidation does not have a detrimental effect on asexual development in Magnaporthe.

A utilizable carbon source is required for conidial germination, whereas continued hyphal growth is dependent on availability of nitrogen (Smith and Grula, 1981). The pathways involved in activation of conidial germination include Ca+2/calmodulin, cAMP/PKA, and RAS/MAP kinase-mediated signalling. In addition, through a genetic screen, malonyl CoA synthetase has been identified as one of the important genes involved in conidial germination in Aspergillus (Osherov and May, 2000). Interestingly, only ech1Δ mutant showed significantly reduced conidial germination when compared to the wild type or fox2Δ strain, suggesting that mitochondrial β-oxidation serves as an exclusive source of energy and/or important metabolic intermediate(s) required for conidial activation in M. oryzae. In addition, melanin deposition was highly reduced in ech1Δ and fox2Δ appressoria. If the phenotypic defects observed during appressorial development could be attributed to the lack of a metabolic intermediate(s) or an energy source, then such defects may be suppressed upon supplementation with an easily utilizable carbon source bypassing the requirement for β-oxidation machinery. Moreover, our previous report shows that supplementation with extra glucose partially remediates appressorial function of host penetration in pth2Δ mutant but not in the pex6Δ strain (Ramos-Pamplona and Naqvi, 2006). Interestingly, supplementation with glucose restored appressorial development (morphology and melanization) in both the ech1Δ and fox2Δ mutants, suggesting that both mitochondrial and peroxisomal metabolism may serve as a major source of energy or metabolic intermediate(s) towards a sustained early pathogenic growth (appressorial development) on the host surface.

In M. oryzae, melanin is composed of polymers of 1,3,6,8-tetrahydroxynaphthalene that is synthesized via polyketide pathway. Malonyl-CoA, which is synthesized from the carboxylation of acetyl-CoA, serves as a starter unit in pentaketide synthesis (Fujii et al., 2000). Both ech1Δ and fox2Δ appressoria showed significant reduction in the melanin deposition, indicating that acetyl-CoA from mitochondrial and peroxisomal β-oxidation likely provides the major pool of precursors for melanin synthesis. However, neither ech1Δ nor fox2Δ appressoria show complete loss of melanin layer indicating that while one of the β-oxidation pathways is defective, the other functional pathway may contribute to acetyl-CoA production. In addition, there are other metabolic pathways, other than β-oxidation cycles, which may also generate acetyl-CoA in eukaryotes. For example, ATP citrate lyase (ACL) produces cytosolic acetyl-CoA in a catalytic reaction involving citrate and coenzyme A (Fatland et al., 2002). In plants, ACL-derived acetyl-CoA has been shown to be carboxylated to malonyl-CoA and then utilized for the synthesis of metabolites essential for development (Fatland et al., 2005). Thus, it is likely that one or more cellular pathways, including mitochondrial and peroxisomal lipid catabolism, operate for melanin biosynthesis during the pathogenic phase in Magnaporthe.

With the help of melanin layer as a semi-permeable barrier between the cell wall and plasma membrane, the appressoria build enormous turgor pressure to subsequently rupture the host by forcing a thin penetration peg through the leaf epidermis (Foster et al., 2003; Talbot, 2003; Wang et al., 2003). As part of the plant defence mechanism, callose is deposited at the entry site of the fungal penetration peg (Jacobs et al., 2003). This plant-derived callose deposition can be visualized by aniline blue staining, and is taken as a measure of fungal penetration peg formation (Vogel and Somerville, 2000). As fatty acid oxidation serves as an important metabolic pathway, its disruption may result in an impairment of the physiology and fitness, which may be reflected in other secondary defects. The fox2Δ strain with significantly reduced melanization is absolutely impaired in host penetration (only ∼2% functional appressoria), and therefore non-pathogenic. Our data are only partly consistent with a previous report in Magnaporthe (Wang et al., 2007), which showed 51% of fox2Δ appressoria to be capable of host entry. It remains to be seen whether such discrepancy is due to the different genetic backgrounds used to generate the fox2Δ mutants in the two studies. When compared to the wild type, around 20% (P = 0.005) of the ech1Δ appressoria could penetrate the host tissue, but were unsuccessful in colonizing the host and led to significantly reduced pathogenesis. Thus, while both peroxisomal and mitochondrial β-oxidation contributes to appressorial melanin synthesis, the former pathway is indispensable for host penetration. Although not essential for host invasion, mitochondrial lipid catabolism is crucial for post-penetration growth and development in Magnaporthe.

In response to penetration by the pathogen, the host starts mounting a defence response, which includes elevation of ROS and induction of genes encoding pathogenesis-related proteins. Success of the pathogen depends on its ability to overcome these elevated ROS levels by employing efficient antioxidant systems. GSH-dependent system, among various antioxidant molecules and enzymes, provides cellular protection from ROS. There are a number of evidences that support the role of intracellular GSH in protecting the cell against ROS under pathological and toxicological conditions (Babson et al., 1981; Meister, 1982; Shan et al., 1990; Gerard-Monnier and Chaudiere, 1996; Smith et al., 1996). Significant protection against paraquat-induced cell damage was achieved in rat alveolar type II cells by exogenously supplemented GSH (Lash et al., 1986). Similarly, supplementation with levocarnitine, which facilitates the transport of long-chain fatty acids across the mitochondrial membrane for β-oxidation and subsequent energy production (Goral, 1998), improved the antioxidant status of the blood and skeletal muscle mitochondria of aged rats (Kumaran et al., 2003). Certain metabolic pathways, while generating key intermediates or by-products, contribute to the redox status of the organelle or cellular milieu. Here, we tested whether any functional relationship exists between peroxisomal (β-oxidation and glyoxylate) and mitochondrial lipid catabolism during pathogenic development in Magnaporthe. Interestingly, appressorial development and/or host penetration defects in peroxisomal mutants could be suppressed by excess glucose [(Ramos-Pamplona and Naqvi, 2006); and the present study] or NAD+ and pyruvate (Bhadauria et al., 2012). However, exogenous NAD+ and pyruvate failed to support invasive growth in mitochondrial ech1Δ mutant. On the other hand, antioxidant agent GSH significantly suppressed host colonization defect in the ech1Δ but not in the fox2Δ mutant. Furthermore, although the glucose-treated pth2Δ penetrated the host cuticle, the resultant penetration hyphae failed to colonize the host tissue (Ramos-Pamplona and Naqvi, 2006). This suggests that Magnaporthe largely depends on energy-generating metabolites during appressorial development and maturation on the nutrient-limiting host surface, and such metabolites are likely not important upon host entry; rather it needs to counter the host defence response that is centred on oxidative burst. It is likely that acetyl-CoA from peroxisomal β-oxidation serves as a precursor for the glyoxylate cycle and indirectly contributes to the conversion of glyoxylate to pyruvate, which is then converted to carbohydrates associated with re-oxidation of NADH to NAD+, whereas acetyl-CoA from mitochondrial β-oxidation may serve as an input for a different pathway such as tricarboxylic acid cycle and as a result may not play an important role in the synthesis of metabolic intermediates like pyruvate and NAD+ via the glyoxylate cycle. Thus, it is likely that peroxisomal lipid catabolism contributes towards the metabolic requirements during appressorial development and host penetration, while the mitochondrial pathway plays a key role in maintaining the redox status during host penetration and colonization by Magnaporthe.

The vital metabolic functions of an organelle may have a role in maintaining its morphology, size and abundance. The mechanism behind this regulation is still not clear. It has been shown in cultured fibroblast cells that deficiency of even single β-oxidation enzyme, including acyl-CoA oxidase (AOX) or D-3-hydroxyacyl-CoA dehydratase/D-3-hydroxyacyl-CoA dehydrogenase bifunctional protein (D-BP), leads to aberrant morphology of the peroxisomes (Funato et al., 2006). Similarly, the loss of multifunctional β-oxidation enzyme Mfe2 in Y. lipolytica resulted in the formation of predominantly larger peroxisomes (Smith et al., 2000). Importantly, fox2Δ strain contained significantly larger peroxisomes in Magnaporthe (Fig. S12). In Drosophila, the deletion of SCULLY, which encodes the third enzyme of the β-oxidation pathway, results in the formation of fewer and smaller mitochondria with reduced and swollen cristae (Torroja et al., 1998). It is worth noting that in the absence of Ech1 function, mitochondria were mostly spherical in morphology when compared to the tubular network in the wild-type Magnaporthe. In addition, the mitochondria in the ech1Δ strain were less dynamic compared to those in wild-type Magnaporthe. Mitochondrial biogenesis is a dynamic process of fission and fusion events that give rise to new mitochondria. A dynamic equilibrium between these two events regulates mitochondrial integrity. Proteins such as Drp1 and Fis1 act as fission proteins, whereas Mitofusin (Mfn1) and Opa1 are involved in fusion process. It has been shown that excessive nitric oxide (NO) enhances mitochondrial fission giving rise to free fragmentation of mitochondria via S-nitrosylation of Drp1 protein (Nakamura and Lipton, 2010). Similarly, overproduction of a misfolded short-chain acyl-CoA dehydrogenase leads to punctate (grain-like) mitochondria due to elevated ROS levels as such fission is alleviated by mitochondria-targeted antioxidant MitoQ (Schmidt et al., 2010). It is likely that the punctate morphology of the mitochondria in the ech1Δ Magnaporthe could be due to aberrant fission of the organelle resulting from enhanced internal ROS levels. Interestingly, we observed that exogenous supply of antioxidants such as GSH/ascorbate restores punctate morphology of the ROS (menadione)-induced damaged mitochondria to normal tubular network (Y.L. He, Y.Z. Deng and N.I. Naqvi, unpublished). At the same time, it could also be possible that the aberrant morphology of the mitochondria in the ech1Δ mutant is due to defect in mitochondrial fusion. Ergosterol is involved in membrane fusion, and mutations in genes including ERG6, ERG24 and ERG28 exhibit morphological defects in mitochondria in budding yeast (Dimmer et al., 2002). Importantly, the ech1Δ mutant shows less ergosterol content in the plasma membrane when compared to that in the wild-type Magnaporthe. In addition, the researchers have shown role of at least 10 different proteins (Mdm30 to Mdm39) in maintaining mitochondrial morphology (Dimmer et al., 2002). Fragmented mitochondria seen in the ech1Δ mutant resemble those observed in budding yeast strains carrying mutations in MDM30, MDM37 or MDM39. In future, it would be interesting to study the mitochondrial fission and/or fusion machinery in the wild type as well as ech1Δ in Magnaporthe.

In conclusion, mitochondrial β-oxidation is not only important for activation of conidial germination, but is also crucial for proper appressorial function and sustained invasive growth of Magnaporthe during establishment of rice blast disease. Although involved in a similar function of β-oxidation of fatty acids, peroxisomal and mitochondrial pathways likely contribute differentially to special metabolic requirements during early and late stages of pathogenic development in Magnaporthe. It remains to be seen whether lipid signalling directly regulates the organellar fission and/or fusion machinery, and consequently the integrity/dynamics of mitochondria and peroxisomes in Magnaporthe.

Experimental procedures

Fungal strains and culture conditions

Magnaporthe oryzae wild-type strain B157 was obtained from the Directorate of Rice Research (Hyderabad, India). Prune agar medium (PA; per litre: 40 ml of prune juice, 2.5 g of lactose, 2.5 g of sucrose, 1 g of yeast extract and 20 g of agar, pH 6.0) was used for standard culture maintenance and conidiation. PA plates inoculated with mycelial plugs were incubated in the dark at 28°C for 3 days and then exposed to continuous light at room temperature for further 5 days to induce conidiation. For harvesting conidia, 10 ml of sterile water (containing 100 μg ml−1 carbenicillin and 100 μg ml−1 streptomycin sulphate) was added to the conidiating colonies and gently scraped with an inoculating loop. The suspension was filtered through two layers of miracloth to remove mycelial debris and conidia were collected by centrifugation. Conidial suspensions were adjusted to the required concentrations after counting with a haemocytometer.

Carbon source utilization was tested on basal medium consisting of 0.67% yeast nitrogen base without amino acids, 0.1% yeast extract, adjusted to pH 6.0 with Na2HPO4, and 2% agar was used. The carbon sources tested were 1% glucose, 1% olive oil with 0.05% Tween20 (as emulsifier), 50 mM sodium acetate, 5 mM valeric acid, hexanoic acid, octanoic acid, 2.5 mM decanoic acid, lauric acid, myristic acid, palmitic acid, or oleic acid. Colony growth was evaluated on plates inoculated with mycelial plugs of wild-type or test strains and incubated at 28°C for 10 days.

For the evaluation of the sensitivity towards ROS, wild type or the ech1Δ was grown on complete medium (CM) containing different concentrations of menadione (25, 50 or 100 μM; Sigma, USA). The growth zone was assessed after 5–7 dpi. Sensitivity towards CFW was evaluated as reported earlier (Ramos-Pamplona and Naqvi, 2006). Effect of NAD+ and pyruvate on invasive growth was assayed as described (Bhadauria et al., 2012). Cytorrhysis assays were performed using established protocols (Sun et al., 2006).

Plasmid constructs and fungal transformants

ECH1 (MGG_12868.6) and FOX2 (MGG_06148.6) gene-deletion mutants were generated using the standard one-step gene replacement strategy in M. oryzae. Briefly, the 1 kb regions immediately upstream and downstream of the ECH1 were PCR amplified (Table S1) and cloned sequentially at HindIII/PstI and BamHI/KpnI sites of pFGL97 (Bar), respectively, to obtain pFGL377. Similarly, the 1 kb regions immediately upstream and downstream of the FOX2 locus were PCR amplified (Table S1) and cloned sequentially into the pFGL44 (Hph) vectors at EcoRI/BamHI and PstI/HindIII sites, respectively, to obtain pFGL428. Plasmid constructs were confirmed by DNA sequencing prior to introduction into the wild-type Magnaporthe strain. For genetic complementation of the ech1Δ strain, the genomic fragment carrying the full-length locus with at least 1 kb of upstream and downstream sequences was PCR amplified and cloned into pFGL44. Plasmid constructs were introduced into Magnaporthe via Agrobacterium-mediated transformation (Mullins et al., 2001). Transformants were analysed by Southern blot analysis to confirm the correct insertion event (Figs S2, S7 and S8).

A short stretch encoding N-terminal 29 residues of Ech1 was fused to the N-terminus of GFP (MTS-GFP). The recombinant plasmid was then transferred to wild-type Magnaporthe. Similarly, to mark the mitochondria, GFP was fused to the C-terminus of MGG_07752.6 encoding ATP1 (Atp1–GFP) and expressed in the wild-type Magnaporthe.

Cytoplasmic GFP was expressed using the MPG1 (constitutively expressed gene encoding Magnaporthe hydrophobin) promoter. In some instances, the Magnaporthe strain expressing the cytoplasmic GFP was used as wild type, while a cytoplasmic GFP harbouring ECH1 deletion was analysed in parallel. Both the strains were confirmed by Southern blot analysis (Figs S7 and S8).

Plasmid construct for Pwl2-RFP (prevents pathogenicity toward weeping lovegrass) expression in M. oryzae was a kind gift from the Valent group in Kansas, USA (Khang et al., 2010) and was transferred into the wild type or ech1Δ using G418 resistance as selection.

RNA extraction and real-time qRT-PCR

For real-time qRT-PCR, RNA was extracted from cultures grown for 24 h in basal medium supplemented with either 1% glucose or 1% rice bran oil. Total RNA was extracted using Trizol reagent (Invitrogen) and further treated by RNase-free DNase (Roche Diagnostics, Germany). Two micrograms of total RNA was reverse transcribed into cDNA with oligo(dT) primer using AMV reverse transcriptase (Roche Diagnostics, Germany) according to the manufacturer's instructions. The qPCR (quantitative real-time PCR) was performed with 20 μl of reaction mixtures containing 1 μl of cDNA, 300 nM each sense and antisense primer, and a SYBR green PCR kit (Applied Biosystems, Foster City, CA, USA). The primers designed to amplify 80 to 120 bp fragments are listed in Table S2. The PCRs were carried out with a thermocycler (7900HT; Applied Biosystems, Foster City, CA, USA). The cycling conditions for PCR amplification were: 95°C for 10 min followed by 40 cycles of denaturation at 95°C for 15 s, annealing, and extension at 60°C for 1 min. Relative mRNA levels were determined after normalization to transcript levels of the TUB (MGG_00604.6) gene.

Pathogenicity tests

The detached barley or rice leaf and rice root infection assays were performed to assess pathogenicity in the mutant strains. Three to five centimetres-long leaf pieces from ∼14-day-old barley or ∼28-day-old rice seedlings were washed for 2 min in 40% ethanol and rinsed several times with sterile distilled water. Such surface-sterilized leaf pieces were then mounted on kinetin agar plates (2 μg ml−1 kinetin, 1% agar). Conidial suspensions were inoculated as 20 μl of water droplets with equivalent conidial concentrations (1000 conidia/droplet). The inoculated leaves were incubated in a 22°C humidified chamber with a 16 h light/8 h dark cycle. Disease symptoms were assessed 5–7 dpi. The root infection assay was performed as described elsewhere (Sesma and Osbourn, 2004; Gupta and Chattoo, 2007). Rice leaf sheath inoculation assays were performed as described by Kankanala et al. (2007).


GFP fluorescence was observed using a Zeiss LSM510 inverted confocal microscope (Carl Zeiss Inc., USA) equipped with a 30 mW argon laser. The objectives used were either a 63× Plan-Apochromat (numerical aperture, 1.4) or a 100× Achromat (n.a. 1.25) oil immersion lens. EGFP was imaged with 488 nm wavelength laser excitation, and a 505–530 nm band pass emission filter, while for RFP imaging, a 543 nm laser and a 560 nm-long pass emission filter was used.

Transmission electron microscopy was performed as described before (Soundararajan et al., 2004).

Staining protocols

Filipin was purchased from Polysciences (Warrington, PA, USA), dissolved in dimethyl sulphoxide (DMSO) and used at a final concentration of 5 μg ml−1. Filipin was added to the assay and cells were observed immediately. Phloxine B staining (10 μM final concentration) was performed on 8-day-old culture of wild type or ech1Δ grown on CM.

To visualize mitochondria, fungal samples were incubated for 15 min in a 100 nM MitoFluor Red 594 (Molecular Probes, Invitrogen, USA) solution prepared in 1× phosphate buffered saline.

Appressorial melanin ghosts were prepared by a slightly modified method described by Wang et al. (1996). Briefly, appressoria from wild type or the mutant on inductive glass surface were washed with 1 M sorbitol in 0.1 M sodium citrate (pH 5.0), and covered with 100 μl of same solution. Zymolyase was added to the appressoria at a concentration of 10 μg ml−1, and the system was incubated for 1 h at 30°C, followed by treatment with 4 M guanidinium isothiocyanate for 30 min at room temperature. Appressoria were then covered with 6 M HCl for 30 min at 100°C (glass coverslips with the appressoria on them were placed in a Petri plate and incubated in an oven set at 100°C). The treated appressoria were carefully washed with water and observed under the microscope using bright field optics.

Statistical analyses

Statistical data involving analysis of conidial germination, appressorial development and function were evaluated by one-way ANOVA (analysis of variance) or the Student's t-test.


We thank Barbara Valent (Kansas State University, USA) for the Pwl2-RFP construct, and the Fungal Patho-Biology group (TLL) for helpful discussions and suggestions. We are grateful to Ouyang Xuezhi (TLL) for electron microscopy (EM) and analysis. This research was carried out using intramural funds from the Temasek Life Sciences Laboratory, Singapore.

Conflict of interest

Authors declare that they do not have any conflict of interest.