Wild-type (WT) strain B157 of M. oryzae, obtained from the Indian Institute of Rice Research, Hyderabad, India, was used for the study. The fungus was grown and maintained on YEG medium (glucose 1 g, yeast extract 0.2 g, H₂O to 100 ml), oatmeal agar (Hi-Media), complete medium (CM) or prune agar (PA; Patkar et al., 2012). The spores were harvested from the biomass grown on PA medium. Photomorphogenesis or asexual development assays were done with fungal cultures grown on PA in the dark at 28°C for 3 days and then under constant illumination at room temperature for further 5 days to induce conidiation.

The full-length YPD1 gene (MGG_07173) was isolated from the B157 strain using gene-specific primers (YPD1F and YPD1R; Supplementary Table 1), and verified by sequencing. A stretch of 525 bp, which covers the second and third exon from the ORF of 1,558 bp, was amplified using the primers (YPD1 OF and YPD1 OR; Supplementary Table 1), and was used for making an RNAi construct. This sequence was analyzed by si-FI software tool (http://labtools.ipk-gatersleben.de/) to predict any off-target effects of siRNA. The RNAi construct was prepared in the pSilent vector (Nakayashiki, 2005), to generate a hairpin loop when expressed in the fungus. The vector pSil-YPD1 was constructed by cloning 525 bp of the YPD1 ORF in the sense and antisense orientations on either side of the cutinase intron in the pSilent vector, under the constitutive TrpC promoter and terminator from Aspergillus sp. The Xba I-cleaved fragment of the construct was then mobilized into pCAMBIA 1,300 at the Xba I site, to obtain the binary construct pSil-YPD1. The plasmid construct was confirmed by sequencing. The construct was used to generate RNAi transformants using Agrobacterium tumefaciens mediated transformation (ATMT). The A. tumefaciens strain LBA4404/pSB1 was first transformed with the binary construct pSil-YPD1 via tri-parental mating using an E. coli strain carrying the helper plasmid pRK2013 (Ditta et al., 1980). The transformed A. tumefaciens strain was then used to transform M. oryzae via ATMT as described (Mullins et al., 2001). Briefly, M. oryzae spores were collected from 8 to 10-day old oatmeal agar or PA plates, and concentration adjusted to 10⁶ spores/ml. The A. tumefaciens strain carrying the binary construct pSil-YPD1 construct was grown at 28°C overnight in Luria-Bertani broth. Overnight-grown A. tumefaciens cells were diluted (OD₆₀₀ 0.15) and grown in the induction medium (IM) for another 6 h. Fungal spores were co-cultivated with Agrobacterium induced with 200 μM acetosyringone (Sigma-Aldrich). The M. oryzae transformants were selected on complete medium (CM) or basal medium (BM) supplemented with the appropriate selection agent such as hygromycin B (Sigma-Aldrich) to a final concentration of 200 μg/ml or BASTA (40 μg/ml). The transformants were maintained as monoconidial isolates, as described, to obtain pure cultures (Gupta and Chattoo, 2007).

Fungal biomass, grown in liquid medium, was frozen in liquid nitrogen. Total RNA was isolated using TRIZOL reagent (Invitrogen Life Technologies). The quality of isolated RNA was checked by electrophoresis on formaldehyde gels and quantified by UV spectrophotometry. Total RNA (5 μg) was used to synthesize the first strand cDNA using M-MuLV reverse transcriptase (New England Biolabs) and random hexamer or oligo dT in a 20 μl reaction system (New England Biolabs). The promoter sequence and transcription start sites (TSS) in the YPD1 locus were predicted using “promoter 2.0 prediction” tool (http://www.cbs.dtu.dk/services/Promoter/). Oligonucleotide primers (T0YPD1F and T1YPD1F; Supplementary Table 1) were designed based on the two potential TSS identified by the aforementioned online tool. The YPD1 transcript variants T0 and T1 were then isolated, cloned and sequenced as follows—WT M. oryzae was grown in duplicate in liquid CM in dark at 28°C for 72 h. At the end of 3-day incubation, one set of the culture was shifted under light for photo-illumination for 6 h. The other set was continued in dark for 6 h. Total RNA was extracted from these two cultures, and the cDNA was amplified using the primers T0YPD1F, T1YPD1F, and T1T0YPD1R (Supplementary Table 1). The amplified cDNA products were cloned at the EcoR V site in pBluescript KS⁺. The cloned cDNA fragments were then sequenced using the M13 primer.

Quantitative real-time RT-PCR (qRT-PCR), to examine the expression pattern of YPD1, was performed using total RNA from 10² conidia/ml geminated and grown for 72 h in liquid CM. qRT-PCR was performed on ABI 7900HT (Applied Biosystems, USA) using SYBR Green I and the requisite primer sets (Supplementary Table 1) for YPD1 and different HK genes. Levels of ENVOY and blue light harvesting protein (BLH) transcripts were analyzed using the TaqMan probes (Applied Biosystems).

Detached leaf infection assay was carried out using leaf blades from the susceptible rice cultivar CO39, and conidial suspension of M. oryzae (1 × 10⁵/ml in 0.2% gelatin). The leaf blades were covered with sterile mira cloth soaked in the aforementioned conidial suspension, and incubated in a humid chamber at 20°C with light/dark cycles (light for 14 h at 25°C and dark for 10 h at 20°C) for 4–7 days until full symptoms became apparent. Rice leaf sheath assay was carried out as described (Patkar et al., 2015). Spores were inoculated on the tender leaf sheath, incubated for about 36 h and observed under the microscope for penetration and invasive hyphae.

The colonies were covered with ~5 ml of sterile distilled water. The spore suspension was obtained by carefully scraping the colonies with a sterile loop, and transferred into sterile falcon tubes. The spore suspensions were vortexed for 10–15 s and filtered through a sterile miracloth, to remove most of the mycelia. The filtered spore suspensions were centrifuged at 13,000 rpm for 4 min, washed and re-suspended in appropriate amount of sterile water. The number of spores was counted using a haemocytometer, and was expressed as conidia per unit area of the colony (cm²) per unit volume of the final suspension (ml).

A 20 μl drop of the conidial suspension (~1 × 10⁴/ml) was spotted on hydrophobic cover slips and left in 80% humid environment at 25°C. Appressorial development (morphology and the frequency of appressorium formation) was analyzed at appropriate time points (16–24 h post inoculation—hpi). Efficiency of appressorium formation was measured in at least three different samples, using more than 100 conidia per assay.

M. oryzae YPD1 cDNA (435 bp) was amplified (see Supplementary Table 1 for primers) and cloned at the N-terminus of a 6x His Tag in pET30(a) for bacterial expression. An E. coli strain BL21 was transformed with the construct and was treated with 1 mM IPTG to induce expression of M. oryzae Ypd1. The fungal Ypd1 produced by E. coli was purified by affinity chromatography using an Ni-NTA resin (Thermo Scientific), concentrated and used for immunization of rabbits for generation of polyclonal antibody. The serum was collected after initial immunization with 300 μg of purified Ypd1 and subsequent two booster doses. The polyclonal antibodies were purified using a Protein A column and the titre was checked by ELISA with purified Ypd1 as the antigen. Western blot analysis was performed to confirm the reactivity of the Anti-Ypd1 antibody against purified fungal Ypd1.

Total protein, for the detection of Hog1 or Ypd1, was extracted from the mycelia of the WT or YPD1 KD transformant M. oryzae grown in liquid CM for 72 h, followed by washing and transfer to fresh liquid CM supplemented with or without a stress element for 6 h. The protein samples were electrophoresed on a 15% SDS-polyacrylamide gel, followed by electro-transfer to a PVDF membrane (Hybond-P; GE Healthcare lifesciences). Hybridization with the Ypd1 antibody was studied using a horseradish peroxidase-conjugated secondary antibody followed by detection with SuperSignal West Pico Chemiluminescence substrate (Thermo Scientific). Phosphorylation of Hog1 was studied using anti-phospho p38 MAPK antibody from Peirce Antibodies (Thermo Scientific). An anti-Hog1 antibody (Santa Cruz Biotechnology) was used to detect Hog1 expression as an internal control.

DAPI (Sigma-Aldrich) staining for nuclei was performed as per the manufacturer's instructions. The immunolocalization of Ypd1 was performed using the Ypd1 antibody (1:200), at different stages of fungal development. The samples (conidiophore, mature conidia and germinating conidia) were fixed with 3.7% formaldehyde for 30 min at room temperature, followed by permeabilization using 0.1% Triton X-100. The fixed samples were further treated as described (Patkar and Chattoo, 2006), using anti-Ypd1 primary antibody and corresponding TRITC-conjugated secondary antibody. The samples were observed under an 100x objective on an epifluorescence (BX51; Olympus) or a 60x objective on a Zeiss LSM 700 confocal microscope. Images were processed using ImageJ or Adobe Photoshop CS2.

Data involving analysis of total conidiation, appressorial development, and relative transcript levels of YPD1 in the WT or YPD1 knock-down (KD) transformants were statistically evaluated by the one-way ANOVA using Tukey's multiple comparison test. Relative transcript levels of ENVOY, BLH and various HKs in the WT or YPD1 KD transformants were analyzed by the two-way ANOVA using the Bonferroni method.

We identified an ortholog of S. cerevisiae YPD1 in the rice-blast fungus (MGG_07173) M. oryzae. Although the sequence homology to the S. cerevisiae Ypd1 was relatively low (44%), the M. oryzae counterpart had the conserved HPt domain with 57% sequence identity. Further, in silico analysis showed that M. oryzae YPD1 had a 1,558 bp-long nucleotide sequence containing two introns, and coded for a protein with 161 amino acid residues. Overall, sequence similarity between M. oryzae Ypd1 and that from yeast and fungi ranged from 44 to 84% (Figures 1A,B). Importantly, the essential residues such as histidine (H83) required for phosphorylation and lysine (K86), glycine (G87), and glutamine (Q102 and Q105) involved in functional folding of HPt in S. cerevisiae, were conserved in M. oryzae Ypd1 (arrows; Figure 1B).

The genome annotation had predicted two transcript variants for YPD1 in M. oryzae (https://www.broadinstitute.org/scientific-community/science/projects/fungal-genome-initiative/magnaporthe-comparative-genomics-proj). A previous study involving deletion of YPD1 in M. oryzae showed that although the protein function was not required for viability, it was essential for conidiation (Jacob et al., 2015). Given that photo-illumination is crucial for conidiation or asexual development in M. oryzae, we studied by RT-PCR expression of YPD1 under both dark and light conditions. Interestingly, we found an additional isoform of YPD1, likely resulting from the use of an alternative TSS upon exposure to light, in M. oryzae. The shorter YPD1 transcript variant (T1; 408 bp) was expressed in the M. oryzae vegetative cultures grown in dark or light, whereas the longer isoform of the gene (T0; 462 bp) was found only in the fungal culture placed under photo-illumination (Figure 1C). Ratiometric analysis of the two isoforms of YPD1 showed that the T1 transcript was relatively more abundant than the T0 in the WT grown under photo-illumination (Figure 1D). Further, we amplified, cloned and sequenced these two YPD1 isoforms (see Section Materials and Methods), to study the nucleotide and amino acid sequences of the differentially expressed transcripts in M. oryzae. The additional 54 bp stretch in the N-terminus of the T0 isoform (Supplementary Figure 1A), coding for 18 amino acids, was analyzed for the putative phosphorylation sites using NetphosK (Blom et al., 1999). A potential casein kinase phosphorylation site was identified in the additional 18 amino acid stretch in the T0 isoform in M. oryzae (Supplementary Figure 1B). Thus, expression of two individual transcript variants under different growth conditions, suggests distinct functions for the two YPD1 isoforms in M. oryzae.

We studied sub-cellular localization of Ypd1 in M. oryzae indirectly by immunolocalization using anti-Ypd1 primary antibody and TRITC-conjugated secondary antibody. The samples from various stages of fungal development, such as conidiophore, mature conidia, and germinating conidia, were co-stained with DAPI to aid visualization of the nuclei. The Ypd1 protein was detected in the cytoplasm of the developing, mature, and germinating conidia (white arrow heads; Figure 2). In addition, punctate aggregates of Ypd1 were seen at the tips of both the conidiophore and germ tubes (white arrows; Figure 2). We infer that Ypd1 likely localizes to the cytoplasm during asexual development in M. oryzae.

We transformed WT strain B157 to generate YPD1 KD transformant (Figure 3A), to study function of the HPt in M. oryzae. Among 123 transformants obtained, 15 showed reduction in YPD1 expression by 25–60% when compared to the WT. Transformants with more than 40% silencing could not survive further monoconidial isolation and repeated subculturing on the hygromycin-containing selection medium. We chose two of the KD transformants, RA6 and RA11 with ~40 and 25% silencing, respectively (Figure 3B; P < 0.005), for further studies. To study the role of Ypd1 in M. oryzae, we determined expression of the Hog1 MAPK in RA6 and RA11 mutants. Western blot analysis showed that while the overall expression of Hog1 was largely unaffected, phosphorylation of the MAPK was significantly reduced upon down-regulation of YPD1 in both RA6 and RA11 KD transformants in M. oryzae (Figure 3C). This indicates that the Hog1 MAPK is likely a downstream target of Ypd1 in M. oryzae.

We further tested expression of light-regulated genes such as those coding for BLH protein (MGG_03002) or the PAS/LOV domain containing Envoy protein (MGG_01041). We found that both the genes were significantly down regulated in the YPD1 KD transformant RA6 when compared to the WT (Figure 3D; P < 0.001). Next, we studied expression of the PAS domain HKs in response to light in M. oryzae. Exposure to light significantly induced expression of all the PAS domain HKs by about 6- to 15-fold in the WT (Figure 3E; P < 0.001). However, importantly, when YPD1 was down-regulated, none of the HKs, except for MGG_02665, showed significant up-regulation even upon photo-illumination (Figure 3E; P < 0.001). It suggests that, except for MGG_02665, the light-inducible expression of PAS-domain HKs is dependent on Ypd1 in M. oryzae. Taken together, these results indicate that the fungal Ypd1 likely controls expression and activation of various PAS-domain HKs and MAPKs, respectively, to regulate diverse light-responsive functions in M. oryzae.

Both the KD transformants RA6 and RA11 showed radial vegetative growth comparable to that of the WT on PA medium (top view; Figure 4A). However, unlike the WT that showed fluffy aerial growth, both the KD transformants formed rather flat colonies on PA medium (transverse section; Figure 4A). Total conidiation was highly reduced in the KD transformants RA6 and RA11 when compared to the WT (Figure 4B). The KD transformants RA6 and RA11 formed only 1.1 × 10² and 4.5 × 10² spores, respectively, as compared to 1.2 × 10⁴ conidia produced by the WT (Figure 4B; P < 0.005). We further assayed infection-related morphogenesis (appressorial development) and the ability to invade plant host, in the KD transformants, to study role of YPD1 in M. oryzae. In an in vitro assay, while 89 ± 4.5% of the WT conidia formed appressoria on an inductive surface, only 50 ± 6.8 and 55 ± 5% of the RA6 and RA11 conidia, respectively, could develop appressoria (Figure 4C; P < 0.0005). In the KD transformants, appressoria were aberrant in morphology and melanization, when compared to the WT (Figure 4C). Most of these KD transformant appressoria failed to penetrate the host cuticle in rice leaf sheath inoculation assay (Figure 4D). The few KD transformant appressoria that did penetrate occasionally failed to elaborate the invasive hyphae and colonize the sheath tissue, when compared to the WT (Figure 4D). Thus, neither RA6 nor RA11 KD transformant could develop any obvious disease lesions on rice leaves, in the plant infection assay (Figure 4E). These results indicate that the failure of proper pathogenic differentiation and appressorial function (host invasion) was likely due to impaired signaling in a pathway downstream of Ypd1 in M. oryzae. Thus, we conclude that Ypd1 function is required for proper asexual development (photomorphogenesis) in the rice-blast fungus M. oryzae, likely via regulation of light-responsive gene expression and pathogenesis.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.