Rice blast, caused by Magnaporthe oryzae, is the most devastating disease in rice worldwide (Ebbole, 2007) and destroys 10-30% of rice production annually (Talbot, 2003). As over half of the global human population depends on rice as staple food (Strange and Scott, 2005), intensive research has been conducted to understand its mechanism of pathogenicity towards the goal of developing effective control strategies for the rice blast.

The infection of rice by M. oryzae is initiated by conidial attachment on host surfaces followed by conidial germination. Via recognition of physical cues on the rice surface, such as hydrophobicity and surface hardness, emerging germ tubes initiate formation of a specialized infection structure, called the appressorium, within 4-6 hours (Hamer and Talbot, 1998; Howard and Valent, 1996). The appressorium is a dome-shaped cell that enables the fungus to penetrate the rice leaf surface. Enormous turgor pressure (8 MPa) builds up inside the melanized appressorium due to high glycerol concentration, which drives penetration of the plant surface. After successful infection of plant tissues, visible symptoms appear rapidly. Over two decades of research has unequivocally established the involvement of several signaling pathways, including the cyclic AMP-dependent (Adachi and Hamer, 1998; Lee and Dean, 1993; Mitchell and Dean, 1995; Xu and Hamer, 1996), calcium-dependent (Choi et al., 2009a, 2009b; Liu and Kolattukudy, 1999; Nguyen et al., 2008), and mitogen-activated protein (MAP) kinase signaling pathways (Hamer and Talbot, 1998; Jeon et al., 2008; Soanes et al., 2012; Thines et al., 2000; Xu and Hamer, 1996) in appressorium differentiation.

One notable advantage that has facilitated studies on appressorium differentiation is the possibility to induce appressorium formation on artificial hydrophobic surface as well as hydrophilic surfaces in the presence of cAMP (Lee and Dean, 1993). Multiple approaches have been deployed to identify genes important in M. oryzae appressorium development, including suppression subtractive hybridization (Lu et al., 2005), super serial analysis of gene expression (Irie et al., 2003; Soanes et al., 2012), expressed sequence tag analysis (Oh et al., 2008), and RNA-sequencing (Soanes et al., 2012). Identification of genes expressed in appressoria via such approaches and subsequent functional analysis of identified genes via mutagenesis have helped understand the molecular mechanism underpinning the initiation and maturation of M. oryzae appressorium. Even though a large number of genes involved in appressorium differentiation have been identified via functional validation, many more genes likely remain to be identified.

In this study, we applied cDNA-amplified fragment length polymorphisms (cDNA-AFLP) analysis (Park et al., 2009) to identify and characterize the function of selected genes using the previously constructed ATMT mutant library (Jeon et al., 2007).

M. oryzae strain KJ201 (wild-type) was obtained from the Centre for Fungal Genetic Resources at Seoul National University, Seoul, Korea. All fungal isolates were grown on V8-juice agar (8% V8 juice, pH 6.8) or oatmeal agar (5% oatmeal [w/v], 2% agar [w/v]) at 25°C under constant light to promote conidial production. The wild-type and ATMT mutants were cultured on complete agar medium, minimal agar medium, C starvation medium, and N starvation medium to observe mycelial growth and colony characteristics (Talbot et al., 1997).

Total RNA was isolated from appressoria, formed on hydrophobic glass for 4-6 hr, and mycelia cultured in complete liquid medium. Extracted RNA (5 μg) was used to produce random primed cDNA with the Prom-II Reverse Transcription System (Promega, Madison, WI, USA). Three independent experiments were performed.

This analysis was performed as previously described (Park et al., 2009). cDNA (200 ng) was digested with EcoRI and MspI and ligated to EcoRI and MspI double strand adapters (Park et al., 2009). The products were diluted (50×) with TE buffer and 5 μl were used for selective amplification, using 36 cycles that include 13 touchdown cycles comprising a reduction of the annealing temperature from 65°C to 56°C in 0.7°C steps, which were subsequently maintained for 23 cycles. A total of 28 primer combinations (Table 1) were used for selective amplification as previously described (Park et al., 2009).

Bands of interest were selected, excised from the gel, and re-amplified as previously described (Park et al., 2009). PCR products were confirmed on 1.0% agarose gels. The amplified transcript-derived fragments (TDFs) were cloned into the plasmid pGEM-T Easy vector system I (Promega). Sequencing of cloned TDFs was conducted on an Applied Biosystems sequencer (ABI3700, Applied Biosystems, Foster City, CA, USA).

Database searches were performed using the BLAST Network Service (NCBI, National Center for Biotechnology Information) and the Magnaporthe genome database (http://www.broad.mit.edu/annotation/fungi/magnaporthe/). To determine whether the genes derived from TDFs were orphan genes or had orthologs in other organisms, the BLAST Matrix program incorporated in CFGP (Park et al., 2008) and InParanoid algorithms (Remm et al., 2001) were used. We applied the cutoff e-value of less than 10^-50 for protein similarity in BLAST Matrix searches and the InParanoid default parameters.

Quantitative real-time RT-PCR (qRT-PCR) was applied to measure transcript levels using specific primers (Table 2) and the conditions described previously (Park et al., 2013). All reactions were performed with three biological and three technical replicates for each sample. The β-tubulin gene was used as the internal control for normalization. All amplification curves were analyzed with a normalized reporter threshold of 0.1 to obtain the threshold cycle (Ct) values. The comparative ΔΔ Ct method was applied to evaluate relative quantities of each amplified sample product. Fold changes were calculated as 2^−ΔΔCt (Livak and Schmittgen, 2001). We applied a fold-change cutoff of ≥1.5 for up-regulation and ≤0.5 for down-regulation.