Introduction

Rice blast disease, caused by the ascomycete fungal pathogen Magnaporthe oryzae (anamorph Pyricularia oryzae) (Couch and Kohn 2002; Kato et al. 2000), is generally considered the most important rice disease worldwide. The disease could result in substantial grain yield loss (Zeigler et al. 1994). The major methods currently used for disease control include fungicides, resistant varieties and cultural practice. The use of fungicides is effective but can cause environmental pollution. While cultural practice can be effective, it has limitations because the specific rice blast virulent isolates that infest in any particular year cannot be predicted. The use of host resistance remains the most cost-effective method for disease management strategy (Hulbert et al. 2001).

Rice blast resistant genes (Pi genes) have been identified and incorporated into rice cultivars for managing rice blast disease throughout the world (Zhou et al. 2007). The Pi genes are effective at protecting against infection of M. oryzae races that contain corresponding avirulence genes (AVR genes) (Silué et al. 1992). Thus far, more than 80 Pi genes for blast resistance have been reported and some of them have been used to control blast disease (Ballini et al. 2008). Eleven Pi genes have been cloned—namely, Pi-ta (Bryan et al. 2000), Pi-b (Miyamoto et al. 1996; Wang et al. 1999), Pi-2/Pi-zt (Zhou et al. 2006), Pi-d2 (Chen et al. 2006), Pi-9 (Qu et al. 2006), Pi-36 (Liu et al. 2007), Pi-37 (Lin et al. 2007), Pikm (Ashikawa et al. 2008), Pi-d3 (Shang et al. 2009), Pi-5 (Lee et al. 2009) and Pi-t (Hayashi and Yoshida 2009). These cloned Pi genes encode the putative cytoplasmic nucleotide binding site and leucine rich repeat (NBS-LRR) proteins with the exception of the Pi-d2 product, which is a putative transmembrane β-lectin-kinase (Chen et al. 2006).

The resistant genes have effectively controlled rice blast disease; however, they usually break down several years after the release of resistant rice cultivars (Lauge and De Wit 1998, Tosa et al. 2005). It has been hypothesized that the ability of AVR genes to defeat R genes is caused by instability and/or high levels of variation of the AVR genes (Khang et al. 2008). To date, 25 AVR genes in M. oryzae have been described (Dioh et al. 2000) and nine AVR genes have been cloned and characterized—namely, AVR-Pita (Orbach et al. 2000), AVR-CO39 (Farman and Leong 1998), PWL1 (Kang et al. 1995), PWL2 (Sweigard 1995), ACE1 (Fudal et al. 2005), AVR-Pizt (Li et al. 2009), AVR-Pia, AVR-Pii, and AVR-Pik/km/kp (Yoshida et al. 2009). The AVR-Pita1 gene, located in the telomeric region of chromosome 3 of M. oryzae genome, encodes for a putative zinc metalloprotease protein (Orbach et al. 2000). The AVR-Pita1 gene product is recognized by the Pi-ta resistant gene product from rice (Jia et al. 2000). The Pi-ta resistant gene has been regularly used in rice breeding programs worldwide to control rice blast disease (Jia et al. 2004; Moldenhauer et al. 1990). It encodes the predicted putative cytoplasmic NBS-LRR protein which triggers the immune response that follows the ‘gene-for-gene' concept. However, epidemics of rice blast disease have occurred on the Pi-ta gene containing rice cultivars, suggesting that it has been defeated (McDowell and Woffenden 2003; Dai et al. 2010). Recently, many studies have shown that the structural variation of AVR-Pita1 alters Pi-ta-specific recognition, for example, a frame-shift mutation in the first exon of AVR-Pita1, which creates a premature stop codon after the 41st amino acid; partial or complete deletion of AVR-Pita1; Pot3 transposon insertion in the coding region corresponding to the AVR-Pita1 protease motif; and base substitutions in the AVR-Pita1 coding sequence (Kang et al. 2001; Zhou et al. 2007; Takahashi et al. 2010). The objective of this present study was to examine the sequence variation of AVR-Pita1 in Thai rice blast fungus isolates. Thirty isolates from northern and northeastern Thailand were collected in 2005 and 2010. The AVR-Pita1 gene from Thai rice blast isolates was cloned and sequenced. Sixty AVR-Pita1 sequences previously reported in the GenBank database (http://www.ncbi.nlm.nih.gov/genbank) were also downloaded and compared. The analysis of the polymorphism patterns based on the DNA sequences, their molecular evolution and the selective forces shaping the evolution of the AVR-Pita1 gene in M. oryzae were reported.

Materials and methods

Infected rice sample collection

Khao Dawk Mali 105 (KDML105) is the most popular Thai aromatic rice variety known by consumers worldwide. KDML105 rice, when cooked, has several distinct characteristics (e.g. highly aromatic, soft and delicious) but KDML105 is susceptible to all major diseases and insect pests including rice blast fungus (Bureau of Rice Research and Development, Rice Department, Thailand 2010). KDML105 is normally used as a susceptibility check variety for the rice blast resistant breeding program in Thailand as KDML105 does not contain any blast resistant genes: Pita, Pib, Pi9, etc. (Srikeaw 2011).

In 2005 and 2010, samples of KDML105 rice variety infected with blast fungus in rice production fields in Thailand were collected for blast fungal isolation. The collection sites were distributed throughout the northern and northeastern regions of Thailand as shown in supplemental Figure S1. Several diseased KDML105 leaf samples from each production field were used for fungal isolation but only one of the single spore isolates from each location was used for Avr-Pita1 cloning, sequencing and sequence analysis.

Fungal isolates and culture

In total, 86 rice blast isolates were collected from the rice variety KDML105 from northern and northeastern Thailand during severe rice blast epidemics in commercial fields in 2005 and 2010 (Table 1). For isolation of single spores, the infected leaves of diseased plants were cut into small pieces and placed on moist filter paper in Petri dishes, then incubated under light for 24 h at 25 °C; single spores were picked with a fine glass needle under a binocular microscope. Each single spore was transferred in to rice flour agar (RFA) medium whose surface was covered with filter paper for 7–14 days. Each isolate was stored at −20 °C on desiccated filter paper and was grown at room temperature under fluorescence lighting to produce mycelia.

Table 1 Accessions of rice blast fungus, host, country of origins and their references
Full size table

DNA preparation

Each rice blast isolate was grown in potato dextrose broth with constant shaking (200 rpm) for approximately 7 days at room temperature to produce mycelia. Fungal mycelia were harvested by filtration through Whatman no.1 filter paper, lyophilized and ground in liquid nitrogen. DNA was extracted from powdered mycelia using cetyltrimethylammonium bromide (CTAB) extraction buffer and incubated at 65 °C for 60 min. The solution was extracted with chloroform/isoamyl alcohol (24:1) and centrifuged at 12,000 rpm at 20 °C for 30 min. After centrifugation, the upper layer was removed to a new tube. The nucleic acid was precipitated by adding the same amount of cold iso-propanol and then incubated at 4 °C for 30 min, centrifuged at 12,000 rpm at 20 °C for 30 min and washed twice with 95 % and 70 % ethanol, respectively. The pellet was dried and dissolved in Tris EDTA buffer. Each DNA sample was quantified using NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA, USA).

PCR amplification and DNA sequencing

Two pairs of primers with overlapping fragments were designed and synthesized based on the genomic DNA sequence of AVR-Pita1 (GenBank ID: AF207841); F1: 5′-AGTGGACCCTTGTCCGATC-3′, F2: 5′-CGCCTTTTATTGGTTTAATTCG-3′, R1: 5′-CCGAAATCGCAACGGTGTG-3′ and R2: 5′-CCTCCATTCCAACACTAACG-3′. These primers were used to amplify and sequence the existence of AVR-Pita1. One primer pair was used to identify the presence and the quality of rice blast fungus genomic DNA; IDMF: 5′-GACCTATGCAATCACCAC-3′ and IDMR: 5′-CGTACTCGAGTGTAATCTCG-3′. This primer was designed from rice blast fungus-specific DNA sequence (GenBank ID: FW343765). All PCR reactions were performed using I-TagTM DNA polymerase (Intron Biotechnology, Seongnam-si, Kyunggi-do, Korea). Each PCR reaction consisted of the following components: 1U of Taq DNA polymerase, 1× Intron PCR buffer, 20 mM MgCl2, 10 mM dNTPs, 1 μL of each 5 μM primer, 30–50 ng of fungal genomic DNA and distilled water in a final reaction volume of 20 μL. Reactions were performed in a GenePro thermal cycler (Bioer Technology, Binjiang, Hangzhou, China) with the following PCR program: one cycle at 94 °C for 2 min for initiation denaturation, followed by 35 cycles at 94 °C for 30 s, 55 °C for 30 s, 72 °C for 50 s, and a final extension of 72 °C for 5 min. The PCR products were separated by 1.0 % agarose gel electrophoresis in 0.5× TBE (Tris-borate-EDTA) buffer, the size of the amplified fragment was estimated using 1 kb Gene RulerTM Express DNA ladder (Fermentas Inc., Glen Burine, MD, USA), stained with ethidium bromide, visualized and photographed using an infinity 3,000 gel photographic system (Vilber Lourmat, Eberhardzell, Germany). PCR products were purified with a Qiaquick gel extraction kit (Qiagen Inc., Valencia, CA, USA) according to the manufacturer’s protocol. After purification, all PCR products were submitted for sequencing by the Pacific Science Co., Ltd. (Thailand).

AVR Pita sequence mining from the GenBank database and sequence analysis

Sixty AVR-Pita1 sequences previously reported in the GenBank database were downloaded. Four sequences from Khang et al. (2008) were referred to as DQ sequences. Eight sequences from Chuma et al. (2011) were referred to as AB sequences. Thirty-eight sequences from Dai et al. (2010) were referred to as FJ sequences and ten sequences from unpublished data were referred to as EU sequences (Table 1 and Fig. 1). DNA sequences of AVR-Pita1 were assembled and aligned by Bioedit software V.7 (Hall 1999) and manually edited by FinchTV V1.4.0 (http://www.geospiza.com/Products/finchtv.shtml). The number of nucleotide diversity per site was estimated as π and θ (Nei 1987; Watterson 1975). Genetic parameters, namely, Tajima’s D test (Tajima 1989), Fu and Li’s D test and Fu and Li’s F test (Fu and Li 1993) and the sliding window analysis, were calculated using DnaSP 5.0 (Rozas et al. 2003). Neutrality and selection tests were performed using DnaSP 5.0. Phylogenetic and molecular evolutionary analyses were conducted using MEGA version 5.0 (Tamura 2011). The phylogenetic trees were constructed using the Maximum likelihood method and drawn using the MEGA program. The stability of tree was evaluated by bootstrap analysis with 1,000 replications.

Fig. 1
figure 1

Geographic distribution of the Magnaporthe oryzae isolates used in the study. Circle (●) represents blast isolates from Thailand (CJ); downward triangle (▼) represents blast isolates from Dai et al. (2010) (FJ); diamond (♦) represents blast isolates from Khang et al. (2008) (DQ); square (■) represents blast isolates from unpublished data (EU) and upward triangle (▲) represents blast isolates from Chuma et al. (2011) (AB) (Table 1)

Full size image

Results

Nucleotide variation of AVR-Pita1

The rice blast resistant gene, Pi-ta, has been effective in preventing infection by races of M. oryzae containing the AVR-Pita1. Eighty-six rice blast isolates were collected from rice variety KDML105 in northern and northeastern Thailand during severe rice blast epidemics in commercial field. Rice blast fungus DNA specific primer was successfully used to verify the present of rice blast fungus genomic DNA. The AVR-Pita1 sequences from 30 blast isolates were successfully amplified using a combination of AVR-Pita1 primers and sequences were deposited in GenBank (GenBank ID: JQ409300–JQ409329). The failure of AVR-Pita1 amplification from 56 blast isolates suggested that DNA sequence at some of these primer sites may have been significantly altered or part of the gene might be deleted. Similar results have been shown by Dai et al. (2010).

Multiple sequence alignment of 30 AVR-Pita1 sequences from Thai blast isolates and 60 previously reported AVR-Pita1 sequences from GenBank (for a total of 90 sequences), were analyzed for nucleotide variation. Sixty-four haplotypes were identified, with 219 segregating sites: 133 in the coding region and 52 in the noncoding region (Table 2). The previously reported AVR-Pita1 sequences from Dai et al. (2010) had the highest number of haplotypes at 37 with 26 of these haplotypes located in the coding region (Table 2). The AVR-Pita1 sequences from Thai blast isolates showed 15 haplotypes with 47 segregating sites; 26 in the coding region and 18 in the noncoding region (Table 2). The length of the complete alignment was 882 bp. Numbering began from the first position of the AVR-Pita1 start codon.

Table 2 Polymorphism and neutral test of different groups of the AVR-Pita1 gene
Full size table

The nucleotide polymorphism (π) of the entire AVR-Pita1 gene was 0.01548 (Table 2). This was caused by the high polymorphism in both coding region (π = 0.01590) and noncoding region (π = 0.01289) (Table 2). The level of nucleotide diversity of the entire AVR-Pita1 gene from previously reported AVR-Pita1 sequences from Chuma et al. (2011) and from Khang et al. (2008) (π = 0.05593 and 0.04261, respectively) was much higher than that in Thai blast isolates reported in this study and those previously reported from Ma et al. (unpublished) and from Dai et al. (2010) (π = 0.00429, 0.00272 and 0.00891, correspondingly) (Table 2).

Tests of neutral selection

Neutrality test of the AVR-Pita1 sequences were examined with three statistical parameters—namely, Tajima’s D, Fu and Li’s D, and Fu and Li’s F (Table 2). The results showed that all statistical parameters were negative. The test values of AVR-Pita1 sequences from Thai blast isolates and from China were significantly deviated from neutrality. In contrast, the test values of previously reported AVR-Pita1 sequences from Khang et al. (2008), Dai et al. (2010) and Chuma et al. (2011) were not significantly deviated from the neutral model (Table 2).

Sliding window analysis was used to characterize the pattern of polymorphism and divergence across the AVR-Pita1 gene. There were substantial nonsynonymous polymorphisms detected in the coding region compared with nonsynonymous divergence in all partitions of AVR-Pita1 protein (Fig. 2). This result was supported by the ratio between nonsynonymous nucleotide polymorphism (πnon) and synonymous nucleotide polymorphism (πsyn) of the AVR-Pita1 sequences (Table 3). The πnon/πsyn ratio of the entire AVR-Pita1 gene and previously reported AVR-Pita1 sequences (AB and FJ sequences) was smaller than 1. On the other hand, the πnon/πsyn ratio of AVR-Pita1 sequences from Thai blast isolates and previously reported AVR-Pita1 sequences (DQ and EU sequences) was greater than 1 (Table 3). The ratio of nucleotide substitutions that lead to amino acid replacements (nonsynonymous substitution, Ka) and nucleotide substitutions that do not lead to amino acid replacement (synonymous substitution, Ks) of all AVR-Pita1 sequences were smaller than 1 except AVR-Pita1 sequences from Thai blast isolates which was much larger than 1 (i.e., 3.995). This extremely high Ka/Ks ratio suggested that the AVR-Pita1 sequences from Thai blast isolates were influenced by adaptive evolution and indicated strong selection for a novel protein function.

Fig. 2
figure 2

Distribution of the AVR-Pita1 allele variation (π value) using sliding window showing three introns and four exons of the AVR-Pita1 gene

Full size image
Table 3 Nonsynonynous over synonymous polymorphism and divergence of the AVR-Pita gene
Full size table

AVR-Pita1 amino acid diversification in blast isolates from Thailand

AVR-Pita1 is known to have three introns and four exons in the open reading frame (ORF) (Orbach et al. 2000). Thirty sequences of AVR-Pita1 from Thai rice blast isolates were translated in to amino acid sequences. The amino acid sequences were aligned and compared with AVR-Pita1 protein of the Chinese isolate O-137 (Orbach et al. 2000). Amino acid alignments were predicted to produce 16 functional proteins including the original AVR-Pita1 protein from O-137 (Table 4). Among these 16 proteins, amino acid variations were predicted to occur at 35 positions including a deletion/insertion. All variations occurred throughout the entire protein (Table 4).

Table 4 Protein variation among 30 AVR-Pita sequences from Thai blast isolates
Full size table

Phylogenetic analysis of AVR-Pita1

The maximum-likelihood method was used to generate a dendrogram, based on Dice’s similarity, for the genetic relationship of 90 AVR-Pita1 sequences (Fig. 3). The phylogenetic analysis revealed two major clades, grouped by country of origin of sequences. The AVR-Pita1 sequences from Thai rice blast isolates mostly clustered together in one clade, separated out from the AVR-Pita1 sequences downloaded from public database, except one isolate (TS 16.2) from Pen, Udonthani which was placed in the other clade. Among 30 blast isolates from Thailand, isolates from 2005 and 2010 are grouped together showing no difference between the times of sample collection. In this clade, it is worth noting that blast isolates from Thailand were placed together with several blast isolates from south and southeast Asian countries including FJ842892 and FJ842890 from the Philippines, AB607338 from Indonesia and FJ842893 and FJ842894 from India. The other clade comprises the AVR-Pita1 sequences mainly originated from China and from the USA. Interestingly, two isolates from Brazil and one isolate from Japan did not group with other AVR-Pita1 sequences in the phylogenetic tree. The result from phylogenetic analysis suggested that blast isolates from Thailand, south and southeast Asia are more closely related to each other and more diverse from rice blast isolates from other parts of the world especially from China and the USA.

Fig. 3
figure 3

Maximum-likelihood tree of the complete DNA sequence of the AVR-Pita1 alleles with bootstrap value of 1,000 replications

Full size image

Discussion

Resistance to M. oryzae in rice follows a gene-for-gene specificity where major resistant R genes are effective in controlling infection by races of M. oryzae possessing corresponding avirulence (AVR) genes (Flor 1971; Correll et al. 2000). The Pi-ta gene is one of the most effective R genes deployed for blast resistance worldwide (Orbach et al. 2000). The effectiveness of the Pi-ta gene relies on the ability to recognize the pathogen’s corresponding avirulence gene AVR-Pita1. The processed AVR-Pita1 protein from the rice blast fungus was demonstrated to interact directly with the translated product of the host R gene Pi-ta in rice triggering resistance (Bryan et al. 2000; Jia et al. 2000). The resistant gene, Pi-ta, is located at 10.6 Mb near the centromere of chromosome 12; a region that often associates with recombination suppression (Bryan et al. 2000). The Pi-ta gene contains two exons interrupted by a single intron and is predicted to be a cytoplastic protein with 928 amino acids with nucleotide-biding site and leucine-rich-repeat domain at the carboxyl terminus (Jia et al. 2009). A single amino acid, alanine at position 918 of the Pi-ta protein, determines its resistance specificity (Bryan et al. 2000; Jia et al. 2000). The genotype variation and resistant/susceptible phenotype at the Pi-ta locus of wild rice (Oryza rufipogon), the ancestor of cultivated rice (O. sativa), was surveyed in 36 locations worldwide to examine the molecular evolution and functional adaption of the Pi-ta gene. The results found that low nucleotide polymorphism of the Pi-ta gene in O. rufipogon was similar to that of O. sativa, but greatly differed from what has been reported for other O. rufipogon genes (Huang et al. 2008). The AVR-Pita1 gene of M. oryzae is located at a telomeric region of chromosome 3 of rice blast fungus (Orbach et al. 2000). The existence of the AVR-Pita1 variants was recently examined in races of M. oryzae from the southern US (Jia et al. 2009). The frequent generation of new virulent races or pathotypes leads to the short life of many newly released blast resistant cultivars (Skamnioti and Gurr 2009; Tharreau et al. 2009). Recent studies on the variations of AVR-Pita1 gene have provided insight into the mechanism of blast genetic variations and instability (Lee et al. 2005). Multiple genetic mutation events and genetics recombination have been found to be the main driving force to creation of new virulent races that overcame major R genes. For example, the study of genetic variation of AVR-Pita1 gene demonstrated that partial deletion, complete deletion, frame-shift mutation and sequence variation have occurred in the AVR-Pita1 sequences among field isolates of M. oryzae from various rice producing countries (Dai et al. 2010). This mention is in agreement with our results which showed the high level of sequence variation of the AVR-Pita1 (Table 2). This result could be explained by the different origins and the different host plant species of blast isolates from Khang et al. (2008) and Chuma et al. (2011), AB and DQ sequences (Table 1).

The statistical tests of neutrality showed the significant negative values of Tajima’s D, Fu and Li’s D and Fu and Li’s F statistics of the entire gene; coding and noncoding regions. These results indicated that the AVR-Pita1 gene possesses diversified sequence structures and is under positive selection pressure in nature. Our results are consistent with previous findings that the population structure of the AVR-Pita1 gene is deviated from neutral model (Kang et al. 2001; Zhou et al. 2007; Dai et al. 2010; Takahashi et al. 2010). Since the statistical tests of neutrality of CJ and EU sequences were significantly deviated from neutrality (Table 2), it suggested that rice blast isolates from Thailand (CJ sequences) and China (EU sequences) are possibly exposed to higher level of selection in nature. This observation was supported by the results of divergence analysis that the πnon/πsyn ratio of CJ and EU sequences was greater than 1 and the Ka/Ks ratio of only CJ sequences (Thai blast isolates) was much larger than 1 (i.e., 3.995) (Table 3). Based on the very high level of nonsynonymous mutation in the coding region of AVR-Pita1 gene, this suggests that the AVR-Pita1 sequences from Thai blast isolates may be influenced by adaptive evolution and indicate strong selection for a novel protein function. Nevertheless, in this study we examined the mutation events at the microspore level of a single spore isolate. The mutations at the microspore level frequently occur in blast fungus and thus the mutation rate obtained from this study might be higher than the actual mutation rate occurring in nature.

KDML105 is a non-glutinous rice variety which is sensitive to photoperiod so it can only be cultivated once a year. KDML105 has been cultivated in the rain-fed areas of northern and northeastern Thailand since 1969. More than 60 cycles of KDML105 crop have been cultivated. In the past three decades rice blast fungus has been the major cause of disease epidemic in Thailand (Smitamana et al. 2000; Rice Department and Thailand 2009). Farmers have been using pesticides for example: isoprothiolane (Fuji- 1 40 % EC), edifenphos (Hinosan 30 % EC) and tricclazole (Beam 75 % WP) to protect their rice production fields (Bureau of Rice Research and Development, Rice Department, Thailand 2010). The long history of KDML105 cultivation and the heavy use of pesticides are corresponding to our finding that the AVR-Pita1 sequences from Thai blast isolates were under the positive selection pressure.

The phylogenetic analysis of the AVR-Pita1 sequences was consistent with the previous report by Khang et al. (2008). This result suggested that Thai blast isolates might have been shaped by an older selective sweep within or near the gene, and other isolates were recently derived from Thai blast isolates (Fig. 3). Our finding suggested that evolutionary mechanism of the AVR-Pita1 gene may have been caused mainly by recurrent selective sweeps.

In the current study, our finding showed high level of nucleotide sequence polymorphisms and the positive genetic selection pressure of the AVR-Pita1 sequences in Thai rice blast isolates. It was also observed by the phylogenetic analysis of the AVR-Pita1 sequences that Thai rice blast isolates were different from blast isolates from other part of the world. The information from this study could draw an attention to the AVR-Pita1 diversification in this part of the world (southeast Asia) where is the origin of species for blast fungus and is rich in genetic diversity of blast fungus isolates.