Characterization and Fine Mapping of a Blast Resistant Gene Pi-jnw1 from the japonica Rice Landrace Jiangnanwan

Rice blast is a destructive disease caused by Magnaporthe oryzae, and it has a large impact on rice production worldwide. Compared with leaf blast resistance, our understanding of panicle blast resistance is limited. The japonica landrace Jiangnanwan from Taihu Lake region in China shows highly resistance to panicle and leaf blast. In this study, three generations (F2:5, F2:6, F2:7) consisting of 221 RILs (recombination inbreeding lines), developed from the cross of Jiangnanwan and Suyunuo, a susceptible-blast japonica variety, were evaluated for panicle blast resistance in the fields and leaf blast resistance in greenhouse in Nanjing in 2013, 2014 and 2015. A blast resistance gene Pi-jnw1 referring to panicle blast resistance and leaf blast resistance was identified in the three generations and located in the region of RM27273 and RM27381 in chromosome 11. The RIL18 line harboring Pi-jnw1 was selected to be backcrossed with Suyunuo to develop BC2F2 populations. According to the genotyping of 1,150 BC2F2 individuals and panicle blast and leaf blast resistance evaluation of 47 recombinants between RM27150 and RM27381, Pi-jnw1 was finally mapped to the 282 kb region between markers W28 and BS39. This study revealed that Jiangnanwan harboring a panicle blast and leaf blast resistance gene Pi-jnw1 could be a genetic source for breeding new rice cultivars with panicle blast resistance.


Introduction
Rice blast, caused by the fungus pathogen Magnaporthe oryzae, is one of the most destructive diseases worldwide, and it occurred in all stages of rice growth [1,2]. The disease pathosystem comprises two major interrelated phases: leaf blast and panicle blast [3]. Compared with leaf blast resistance, less is known about the genetic components for panicle blast resistance, which is indispensable for stable rice production. Leaf blast resistant cultivars may be susceptible to panicle blast, and it implies that the genetic mechanisms of blast resistance might vary across the plant growth stages [4][5][6][7]. The technical problems as lacking of standard inoculation and evaluation systems, variations in heading date and weather conditions, are obstacles to the PLOS ONE | DOI: 10 [3,8]. It encodes a NBS-LRR protein, and can protect WRKY45 from degradation by ubiquitin proteasome system. The blast resistance of cultivar usually can be lost after few years for the genetic instability and pathogenic variability of M. oryzae [9]. Therefore, to further explore new resistance genes especially panicle blast resistance genes from rice landraces will be the most useful strategy in rice blast resistance breeding. Up to date, approximately 100 blast resistance loci or genes have been mapped on 12 chromosomes except chromosome 3 [6,7]. Twenty five blast resistance genes have been cloned [10], and eight of them located in two gene clusters, including three genes Pi2, Pi9 and Piz-t in Pi2 locus and five genes Pik, Pik-m, Pik-p, Pi1 and Pi-ke in Pik locus [11][12][13][14][15][16][17]. Among 25 cloned genes, 23 genes encode NBS-LRR (nucleotide-binding site -leucine-rich repeat) proteins, except Pi21 encodes proline-containing protein and Pid2 encodes receptor kinase [18][19][20]. It has been shown that at least six R genes, Pi1 [21], Pi2 [22], Pi9 [23], Pi5 [24], Pi33 [25] and Pigm [26], probably confer broad-spectrum resistance to a number of isolates from different countries respectively. For instance, Pi9 located on the same region with Pi2, showed resistance to 43 isolates from 13 countries [23]. Pi5, a locus associated with resistance to at least 6 blast races from Philippines and 26 isolates from Korea [24], and Pi33 was resistance to more than 2,000 isolates originating from 55 countries [25].
In our previous research, Jiangnanwan, a japonica rice landrace from Taihu Lake region of China, exhibited broad-spectrum resistance to rice blast [27]. Li et al. [28] concluded that two effect genes might be involved in the leaf blast resistance with F 2 population deriving from a across between Jiangnanwan and a blast-susceptible variety Suyunuo. In this study, we obtained 221 F 2:7 RILs with three generations (F 2:5 , F 2:6 and F 2:7 ), and identified panicle blast and leaf blast resistance genes to the strain Hoku 1 with QTL mapping method. Furthermore, we examine the correlation between the resistance of panicle and leaf blast, and fine mapped the blast resistance gene Pi-jnw1.

Plant materials and growth
Jiangnanwan is a japonica rice landrace from Taihu Lake region in China and has broad spectrum resistance to leaf blast. Suyunuo is a susceptible japonica rice landrace from Taihu Lake region. We developed an F 2 population from a cross between Jiangnanwan and Suyunuo, and three generations (F 2:5 , F 2:6 and F 2:7 ) of 221 recombination inbreeding lines (RILs) were generated by a single-seed descent method (Fig 1).
The populations of F 2:5 , F 2:6 , F 2:7 and two parents (Jiangnanwan and Suyunuo) were evaluated for panicle blast resistance in 2013, 2014 and 2015 at Jiangpu Experimental Station of Nanjing Agricultural University (Jiangsu Province, China; 118˚50 0 E, 32˚02 0 N). Twenty plants of each RIL grew in two rows per test plot, and the spaces were 30 cm between rows and 10 cm between plants within rows. Suyunuo and Jiangnanwan were planted adjacent to the test rows as susceptible and resistant controls respectively. Field management was carried out in accordance with the local production process [29].
The populations of F 2:5 , F 2:6 , F 2:7 and two parents (Jiangnanwan and Suyunuo) were also evaluated for leaf blast resistance in 2013, 2014 and 2015 in the greenhouse. The seeds were sown in plastic trays of 60 × 30 × 5 cm with sieved garden soil as described by Wang et al. [27]. Thirty lines and two parents were sown in each tray, and 6-8 seeds per line were sown for inoculation. Seedlings were grown in greenhouse at 22-30˚C with a light and dark cycle of 16 h and 8 h until they were at the four-leaf stage for disease evaluation.

Inoculation and disease evaluation
To evaluate the panicle blast resistance in the field, 221 RILs (F 2:5 , F 2:6 , F 2:7 ), forty-seven BC 2 F 2 recombinants and two parents at the mid-booting stage were inoculated with the strain Hoku 1 of M. oryzae by the injecting method as described as Liu et al. [30]. Fifteen booting panicles of each line were injected by 1-2 ml blast isolate Hoku 1 conidial suspension (5×10 4 conidia/ml). Three weeks after inoculation, phenotypic evaluation was conducted based on visual assessment of diseased grains percentage as described by Koizumi et al. [31] and the scores were ranged from 0 (without diseased grain) to 100% (100% diseased grains).
Four-leaf stage rice seedlings of Jiangnanwan, Suyunuo, 221 RILs (F 2:5 , F 2:6 , F 2:7 ), and fortyseven BC 2 F 2 recombinants were inoculated with the strain Hoku1 spore suspension (5×10 4 conidia/ml) in inoculation chambers as the method described by Wang et al. [27]. After inoculation, the plants were kept in dark at 26˚C with relative humidity 95% for 24 h, and then transferred to a greenhouse with 25-28˚C and 100% relative humidity by intermittently Characterization and Fine Mapping of a Blast Resistant Gene Pi-jnw1 in Rice Landrace Jiangnanwan spraying water for 2 min every three hours. After seven days of inoculation, lesion scores of 0 to 5 were investigated based on lesion type with appropriate reference of the disease area of each plant as described by Shi et al. [32]. Each line was inoculated with three replications in each experiment and three independent experiments were conducted either leaf blast or panicle blast resistance evaluation.

Genetic map construction and identification
221 RILs of F 2:6 population were used for genotyping and constructing molecular linkage map 0.2 to 0.5 g of leaves at the four-leaf stage from each line of RILs (F 2:6 ) and parents were collected specifically for DNA extraction by using the CTAB method [33]. 2,300 SSR markers kept in our lab and 108 newly designed InDel markers distributed on 12 chromosomes were screened for polymorphisms. 93 markers with polymorphisms between the two parents were finally used for genetic map construction.
All of the PCR reactions with the markers used a 10 μl reaction mixture containing of 1 μl template DNA, 0.5 μl of each primer, 0.1 μl of Taq (0.01U/μl), 1.6 μl of 10×Buffer, 0.2 μl of dNTP and 6.1 μl of ddH 2 O. PCR procedures were conducted as follows: Preheating for 5 min at 95˚C, 32 cycles (40 sec at 95˚C, 40 sec annealing temperature, and 40 sec at 72˚C), finally 72˚C for 10 min. The PCR products were analyzed on the 8% acrylamide gels.
In order to identify panicle blast and leaf blast resistance genes, QTL mapping method was performed using software IciMapping v4.0 (http://www.isbreeding.net/). The software was set LOD > 2.5 as a threshold which must be operate 1000 times at the p< 0.05 level. In this study, the panicle blast resistance QTLs were named as qPbj-A-B, and the leaf blast resistance QTLs were named as qLbj-A-B, in which A means the chromosome number and B means the sequence of QTL.

Data analysis
Experimental data were analyzed using the IBM SPSS Statistics software 19.0, and bivariate analysis method were used for analyzing the correlation between the panicle blast resistance and leaf blast resistance [33].
The RIL18 line harboring Pi-jnw1 was selected from F 2:6 RIL populations and backcrossed with Suyunuo for developing fine mapping populations,.1,150 plants of BC 2 F 2 population generated by 26 resistant BC 2 F 1 individuals were used for constructing fine genetic linkage map and identifying recombinants in the target region of Pi-jnw1. The franking markers RM27150 and RM27381 were used to genotype the 1,150 BC 2 F 2 segregating plants and 47 recombinants were detected. Then the 47 recombinants were inoculated to identify the panicle blast resistance phenotypes and the next generation seeds of 47 recombinants were inoculated to identify the leaf blast resistance phenotypes.

Characterization of resistance to panicle and leaf blast in Jiangnanwan
In 2013, 2014 and 2015, Jiangnanwan and Suyunuo were inoculated with the strain Hoku1 in field for panicle blast resistance evaluation and in greenhouse for leaf blast resistance evaluation. The results showed that Jiangnanwan was highly resistance to panicle blast and leaf blast, while Suyunuo was highly susceptible (Fig 2A-2C, Table 1). The frequency distributions of panicle blast and leaf blast resistance in 221 RILs (F 2:6 , F 2:7 and F 2:8 ) were asymmetric and continuous, and they were all predisposed resistance-inclined distribution (Fig 3A-3G). Similar results of frequency distributions of panicle blast and leaf blast resistance were obtained by the IBM SPSS Statistics software 19.0, and the characteristic parameters (Skewness and Kurtosis) showed the frequency distributions of panicle blast and leaf blast resistance in 221 RILs were all predisposed resistance-inclined distribution (Table 1).  The frequency distributions of panicle blast resistance in three generations of 221 RILs (F 2:5 , F 2:6 and F 2:7 ) were asymmetric and continuous, and they were all predisposed resistanceinclined distribution (Fig 2A-2C). The frequency distributions in the three tested generations under the experimental paddy field and greenhouse were not bimodal, suggesting that multiple loci are involved in the panicle blast and leaf blast resistance of Jiangnanwan (Fig 2D-2F).

Identification of Pi-jnw1
With a linkage map which covering 1,690.76 cM on the 12 chromosomes and an average distance 18.18 cM between two connected markers, a blast resistance gene Pi-jnw1 referring to panicle blast and leaf blast resistance was detected in the same region of RM27273 and RM27381 on chromosome 11 by QTL IciMapping 4.0 in three generations (F 2:5 , F 2:6 and F 2:7 ) ( Table 2, Fig 4).
To determine whether leaf blast resistance was related with panicle blast resistance in Jiangnanwan, the correlation of panicle blast resistance and leaf blast resistance of 221 F 2:5 , F 2:6 , F 2:7 RILs was examined. The results showed that the correlation coefficients between panicle blast  resistance and leaf blast resistance were 0.49 in the F 2:5 RILs, 0.371 in the F 2:6 RILs, and 0.551 in the F 2:7 RILs respectively, all with a significantly positive relationship (Table 3).

Discussion
Panicle blast usually caused more loss of yielding than leaf blast in rice production. However, fewer genetic analyses of rice panicle blast resistance have been reported compared with leaf blast resistance. More field works, complex phenotype evaluation system and the influence of environmental conditions are obstacles to study rice panicle blast resistance. There are various ways for evaluating the panicle blast resistance: (1) Injecting inoculation, injecting 2-3 ml of spore suspension into one booting panicle [30]; (2) Inducing inoculation, controlling the field conditions to be suitable for development and epidemic of blast disease [3]; (3) In vitro inoculation, 6 cm rice panicle necks containing 1-3 rachis nodes were put on the filter paper in petri dishes, then the nodes were inoculated with 2 ml spore suspension by a micropipette [5]. The resistance genes Pi-64 and Pi-25 were identified by vitro inoculation [35,36], and Pb1 was identified under suitable field conditions for blast disease development [3]. We use the modified injecting method to inoculate the 221 RILs and forty-seven BC 2 F 2 recombinants in the field. In our study, the frequency distributions of diseased grains percentages in the three tested generations in different fields were relatively stable and a blast resistance gene Pi-jnw1 referring to panicle blast resistance in chromosome 11 were detected in three years and two minor resistant loci (qPbj-7-1 and qPbj-7-2) in chromosome 7 were both detected in two years. It indicated that the injecting inoculation method might be more appropriate for identifying the panicle blast resistance. Jiangnanwan, one japonica rice landrace from Taihu Lake region in China, exhibited broad-spectrum resistance to leaf blast and highly resistance to panicle blast [37]. Li et al. [28] studied the genetic patterns of leaf blast resistance to Hoku1 in Jiangnanwan with P 1 , P 2 , F 1 and F 2 population deriving from a across between Jiangnanwan and a blast-susceptible variety Suyunuo and concluded that two genes might be involved. In our results, only one blast resistance gene Pi-jnw1 could be detected in 221 F 2:5 , F 2:6 , F 2:7 RILs, respectively. The possible reason could be due to the different populations and different analysis methods. So far, more than 20 blast resistance genes were reported on rice chromosome 11, four of them locate near Pi-jnw1 region. The Pb1 locus was mapped in the Modan-derived chromosomal region in the middle part of the long arm of chromosome 11, located closet with the RFLP marker of S723 [3]. The rice blast resistance gene Pik, which confers high and stable resistance to many Chinese rice blast isolates, encoded two coiled-coil nucleotide binding site leucine-rich repeat (NBS-LRR) proteins [17]. The Pi34 locus was located in the 54.1 kb region on the genomic sequence of Nipponbare and acted partial resistance to blast in Chubu 32 [38]. Pi-hk1 was identified on chromosome 11 of Heikezijing, located between the SSR markers of RM7654 and RM27381 [20]. According to the fine mapping results, Pi-jnw1 was not in the same region of Pb1. In our further study, we will confirm the fine mapping results and use more markers to detect whether Jiangnanwan harbors Pik, Pi34 and Pi-hk1 genes in the Pi-jnw1 region.
In our study, the blast resistance gene Pi-jnw1 was identified both in panicle blast resistance and leaf blast resistance of the three generations (F 2:5 , F 2:6 and F 2:7 ), suggesting that there was a positive relationship between panicle blast and leaf blast resistance detected in Jiangnanwan. It is consistent with the common viewpoint that panicle blast resistance is correlated with leaf blast resistance in many rice cultivars [39]. However, there were also four minor panicle blast resistance specific loci qPbj-6-1, qPbj-7-1, qPbj-7-2 and qPbj-9-1, and it indicates that some loci might be only influence the panicle blast resistance. Interestingly, qPbj-7-1, qPbj-7-2 and qPbj-6-1 were contributed by Suyunuo indicated that there were some genes in Suyunuo against panicle blast which could be detected under specific conditions. In this study, 93 genetic markers with polymorphisms between two parents were used for genetic map construction, and the frequency of polymorphisms between Jiangnanwan and Suyunuo was not as high as the indica/japonica crosses. In further research, more genetic markers between W28 and BS39 will be designed and larger segregation populations will be constructed for fine mapping the Pi-jnw1. The recombinants harboring Pi-jnw1 will be further used for breeding new cultivars with back crossing with the elite cultivars and marker associated selection method (MAS).
Breeding new blast resistant cultivars is considered as an effective and economical way to control this disease. However, among the cloned 25 resistance genes, 24 of them were referring to the leaf blast resistance and few of them have been widely applied in rice breeding. Many cultivars show different levels of partial resistance to leaf and panicle blast. This implies that the genetic mechanisms of host resistance might vary across growth stages. In this study, Jiangnanwan showed broad spectrum resistance to leaf blast and highly resistance to panicle blast, and the panicle blast resistance showed a positive correlation with leaf blast resistance. The mapping results also showed that Pi-jnw1 could be detected with panicle blast resistance phenotypic data and leaf blast resistance phenotypic data in three years and located in the same region of RM27273 and RM27381 on chromosome 11 in the three generations (F 2:5 , F 2:6 and F 2:7 ). It indicated that Jiangnanwan might be a good resource for application in rice breeding programs, and further cloning and functional analysis of Pi-jnw1 could be necessary for clarifying the molecular basis of panicle blast resistance and leaf blast.