Introduction

Bacterial leaf blight (BB) of rice caused by Xanthomonas oryzae pv. oryzae (Xoo) is a devastating disease of tropical and temperate rice. The first report of rice BB was by Japanese farmers in 1884; however, it was not reported to cause serious damage to the rice crop until the appearance of new high-yielding varieties in the 1960s and 1970s (Mizukami and Wakimoto 1969; Mew et al. 1992; Adhikari et al. 1994). Since then, the incidence of BB has been widely reported throughout Asia and BB has become one of the three serious rice diseases causing significant yield loss annually (Chen et al. 2011). Estimates of average yield losses are difficult to obtain because of the extensive geographic and seasonal variation in BB incidence. However, during disease severity, BB has been reported to cause yield losses of up to 80 % (Mew et al. 1993; Srinivasan and Gnanamanickam 2005). Enhancing host resistance is considered as the most effective strategy to achieve disease resistance in rice. However, the high degree of pathogenic variation in Xoo often causes the breakdown of resistance (Vera Cruz et al. 2000; Suh et al. 2013). Expanding genetic sources with novel resistance genes from wild relatives of rice, their deployment in breeding programs, and pyramiding two or more effective resistance genes are some of the approaches envisioned in developing rice cultivars with durable BB resistance to Xoo (Suh et al. 2009; Natraj Kumar et al. 2012). The race-specific interaction between rice and Xoo is thought to follow the classic gene-for-gene relationship (Flor 1971) and is similar to the host–pathogen model for understanding the molecular mechanisms in host resistance to pathogens (Dai et al. 2007). However, BB resistance mechanisms appear to have distinctive differences from other characterized R-genes. Although most of the reported R-genes in other crop pathogen systems are dominant in nature, almost one-third of the R-genes conferring resistance to Xoo have been reported as recessive (Verdier et al. 2012). Comparing the prevalent class of functionally defined R-genes that encode intracellular nucleotide-binding/leucine-rich repeat (NBS–LRR) proteins as immune receptors to initiate defense signaling, the R-genes to Xoo encode various types of proteins (Belkhadir et al. 2004; McHale et al. 2006; Xiang et al. 2006; Hammond-Kosack and Kanyuka 2007). To date, more than 39 R-genes conferring host resistance to various strains of Xoo have been identified and some of those have been characterized (Cheema et al. 2008; Wang et al. 2009; Guo et al. 2010; Miao et al. 2010; Verdier et al. 2012; Wang et al. 2014a; Zhang et al. 2014). A total of 14 recessive genes—xa5, xa8, xa13, xa15, xa19, xa20, xa24, xa25, xa26b, xa28, xa31, xa32, xa33,and xa34—in the series from Xa1 to Xa39 have been identified with the remainder dominant (Chen et al. 2011; Liu et al. 2011). Of these 39 R-genes, nine R-genes—Xa1, Xa3/Xa26, xa5, xa13, Xa10, Xa21, Xa23, xa25 and Xa27—have been isolated and characterized as encoding five types of proteins, suggesting multiple mechanisms of R-gene-mediated Xoo resistance (Song et al. 1995; Yoshimura et al. 1998; Iyer and McCouch 2004; Sun et al. 2004; Gu et al. 2005; Chu et al. 2006; Tian et al. 2014; Wang et al. 2014b). Seven R-genes (Xa4, Xa7, Xa22, Xa30, Xa31, Xa33, and xa34) have been fine-mapped based on morphological and molecular markers. In some of them, however, there are discrepancies such as duplicate nomenclature of genes for the same gene without compelling evidence, for example, for the dominant R-genes Xa25, Xa26, Xa32 and Xa33,and the recessive R-genes xa25, xa26, xa32, and xa33 (Gao et al. 2001; Lee et al. 2003; Sun et al. 2004; Ruan et al. 2008; Korinsak et al. 2009; Zheng et al. 2009; Liu et al. 2011; Kumar et al. 2012). The R-genes are distributed among nine rice chromosomes and 12 of the R-genes are intensively clustered on chromosome 4 (Xa1, Xa2, Xa12, Xa14, Xa30, and Xa31) and chromosome 11 (Xa3/26, Xa4, Xa10, Xa21, Xa22, and Xa23), respectively.

In Korea, BB isolates were grouped into five races (K1–K5) based on the response of five rice cultivars to Xoo infection as a differential system, before the occurrence of the new BB race (K3a) in the southwestern coastal plain area in 2003 (Noh et al. 2003). Approximately 95 % of Korean BB isolates are concentrated in the four main races, K1, K2, K3, and K3a, and the races K4 and K5 have been less frequently reported (Noh et al. 2003; Jeung et al. 2006; Noh et al. 2013). In particular, the K3a race has been reported as the prevailing race in Korea because of fewer resistant cultivars (Jeung et al. 2006).

A series of japonica breeding lines, IR83261-3-7-23-6-2-1-1-2-1-2 (11325), IR83261-5-13-8-4-1-2-2-1-1-1 (11327), and IR83261-5-13-8-4-1-2-2-1-2-1 (11328), derived from indica donor line IR65482-7-216-1-2 were found to have high levels of resistance to the new Xoo race K3a in Korea. The present study was carried out to identify the genetic basis of the new BB R-gene conferring resistance to Korean Xoo race K3a in IR65482-7-216-1-2, to construct a physical map containing the region, and to identify candidate genes. The expression levels of candidate genes were confirmed by RT-PCR for isolation and characterization of this gene.

Materials and methods

Plant materials

Anmi, a japonica rice cultivar, was developed by backcrossing a donor line (IR65482-7-216-1-2) to the recurrent parent Junam. Anmi demonstrated a susceptible reaction to the Korean Xoo strain K3a isolate (HB01009). However, in contrast to Anmi, two of its sister lines (advanced backcrossed lines 11325: BC3F8 and 11327: BC4F8) demonstrated resistance to K3a (Supplementary Fig. 1). The F1 progenies from crosses between 11325 and the highly susceptible cultivars Anmi and Ilpum were used to examine whether the BB resistance gene acted as a dominant or recessive trait. Two F2 populations composed of 787 and 796 F2 plants were created from the crosses Anmi/11325 and Ilpum/11325, respectively, for physical mapping of the target gene. Each F2 individual was self-fertilized to obtain a set of F3 lines. Another F2 population, derived from a cross between IRBB4 (Oryza sativa sub sp. indica) carrying Xa4 and 11325 carrying the new gene, was used for an allelism test between Xa4 and the target R-gene. The cultivar IR24, one of the progenitors of IR17494-32-3-1-1-3, and the near-isogenic lines IRBB1, IRBB3, IRBB5, IRBB8, IRBB10, IRBB13, IRBB21, and IRBB57 were used in haplotype analysis. Seeds of these lines were obtained from the Plant Breeding, Genetics, and Biotechnology Division of the International Rice Research Institute, Los Baños, Laguna, Philippines.

BB inoculation and evaluation of resistance

Four different virulent strains—K1(HP01013), K2(HP01014), K3(HP01015), and K3a(HP01009)—of Xoo from Korea were used for the BB evaluation. At maximum tillering stage, the F1 progenies, the F2 individuals, and F3 lines derived from crosses 11325 (11327) and Anmi, Ilpum, and IRBB4 were inoculated with the strains using the leaf-clipping method (Kauffman et al. 1973) under field conditions in Suwon, Korea. Reaction to the pathogen was evaluated at 14 days after inoculation. Evaluation of leaf damage level by the pathogen was carried out by measuring the average lesion length of three leaves (highly resistant <1 cm, resistant 1–3 cm, moderately resistant 3–5 cm, susceptible 5–10 cm, highly susceptible >10 cm).

PCR conditions

Cultivar Anmi, the two sister lines (11325 and 11327), and the resistant donor (IR65482-7-216-1-2) line were used to determine the location of the new resistance gene conferring resistance to K3a. The lines were compared through PCR analysis using a total of 287 DNA markers (SSR 221 and STS 66) selected for an even coverage of the rice genome. PCR amplification was performed in 20-μl volumes of reaction mixture containing 10 pMol of each primer, 50 mM KCL, 1.5 mM MgCl2, 25 ng template DNA, and 0.02 U/μl of GenDepot DNA polymerase. The PCR condition used consisted of one cycle of 8 min at 95 °C, followed by 35 cycles of 30 s at 55 °C, 1 min at 72 °C, and a final cycle of 10 min at 72 °C.

Development of DNA markers

To detect co-dominant DNA markers with the R-gene, additional DNA markers were developed in the region of the flanking markers RM1233 and RM5766. Primer sequences of published rice microsatellite markers (RM) located within the region were obtained from the Gramene database (http://www.gramene.org/; McCouch et al. 2002). Markers for detecting insertion-deletion (Indel) polymorphisms were developed using the DNA polymorphism database (http://shenghuan.shnu.edu.cn/Default.aspx?tabid=5641; Shen et al. 2004). Cleaved amplified polymorphic sequence (CAPS) markers were applied to construct high-resolution fine-mapping and haplotype tests using the sequencing data of gene-based STS-PCR products. Genomic DNA sequences for each bacterial artificial chromosome (BAC) clone were determined through primer walking based on the corresponding sequences in the Nipponbare genome by GreenGene Bio Tech, Inc. (Seoul, Korea). To design primers for Indel markers, Primer3 version 4.0 (http://frodo.wi.mit.edu/primer3/) was used (Rozen and Skaletsky 2000). Four gene-specific DNA markers, MP1+MP2, 10571.T17/Hinf1, and U1/I1, tightly linked to the resistance genes Xa4, Xa3, and Xa21, respectively, were used to analyze the gene validation.

Graphical mapping/gene annotation

F2 populations were used to determine the exact position of the nearest recombination event to the target region on chromosome 11. The genetic fine-mapping of the target R-gene was constructed according to the physical distance of the DNA markers on BAC and P1-derived artificial chromosome (PAC) clones of cv. Nipponbare released by the International Rice Genome Sequencing Project (IRGSP). All genes with clear open-reading frames (ORFs) were analyzed based on the available rice genome sequence and annotation databases from NCBI (www.ncbi.nim.nih.org/unigene) and TIGR release 7.1 (http://rice.plantbiology.msu.edu/) and a putative function for each gene identified in the region of interest was annotated using BLAST-P utility (www.ncbi.nlm.nih.gov).

cDNA synthesis and semi-quantitative reverse transcription PCR (RT-PCR)

Total RNA was extracted from collected rice samples by using TRIzol reagent (Invitrogen, UK) according to the manufacturer’s protocol. DNA was removed from RNA samples by using TURBO DNA-Free DNase (Ambion, Inc.). Total RNA was quantified by measuring absorbance at 260 and 280 nm using the Nanodrop ND 1000-spectrophotometer (Thermo Fisher Scientific, Inc., Wilmington, NC, USA).

Total RNA from leaves collected at 1, 4, 8, 24, and 48 h after inoculation was used to synthesize each pool of cDNA using the SuperScript III First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA). Briefly, cDNA synthesis from 10 μg of total RNA is incubated with 2 µl oligo (dT)18 primer. Subsequently, the buffer, dNTPs (final conc. 1 mM), 400 U M-MuLV Reverse Transcriptase, and 40 U RiboLock™ RNase Inhibitor were added and incubated for 60 min at 42 °C. The reaction was terminated through heating at 70 °C for 5 min. Semi-quantitative RT-PCR analyses were performed with 250 ng of cDNA for 28–35 cycles, with an annealing temperature of 55–62 °C. Rice actin1 (Osactin) was used as an internal control, with the forward primer 5′-CTGCTATGTACGTCGCCATC-3′ and the reverse primer 5′-AGTCTCATGGATACCCGCAG-3′, and the same PCR conditions as described above.

Results

Resistance reaction to Korean BB races

The breeding lines and cultivars tested in this study were estimated for resistance levels against four BB isolates representing the spectrum of virulence in Korea (Table 1). All the plants, including International Rice Bacterial Blight (IRBB) NIL IRBB3, showed a resistant reaction to K1(HB01013), K2(HB01014), and K3(HB01015) isolates. However, in the case of K3a(HB01009), only three lines—11325, 11327, and IR65482-7-216-1-2 (containing the unknown BB R-gene)—showed strong resistance (Table 1).

Table 1 Leaf reaction of tested plants to four different BB isolates
Full size table

Detection of substituted chromosomal segments

To ascertain the donor-derived segments conferring resistance to BB race K3a, a total of 261 DNA markers were used for background selection of Anmi along with two advanced backcross lines (ABLs), 11325 and 11327, and a graphical map constructed to determine the genomic position of the new BB R-gene, tentatively designated as Xa40(t) (Fig. 1). The substituted segments were distributed on chromosomal regions in the genetic background of two ABLs. On the graphical map, line 11325 had 10 substituted segments on chromosomes 1, 2, 6, 8, 11, and 12, whereas line 11327 had two substituted segments on chromosomes 4 and 11, respectively (Fig. 1). Substituted segments were not found on the remaining chromosomes in this genetic background survey. In the validation test with the F2 population, including the new gene, we could exclude other candidate regions to narrow down the target region by fine-mapping. The analysis of introgression fragments positively showed the target region of Xa40(t) to be the telomeric region on the long arm of chromosome 11, as this section overlapped in the two R-lines, flanked by two polymorphic markers, RM 1233 and RM 5766.

Fig. 1
figure 1

Background selection of two tested lines in Anmi genetic background. A is Anmi, D1 and D2 are advanced backcross lines 11325 and 11327, respectively. Blue bars indicating introgressed DNA fragment from the R-donor plant show the polymorphic region between Anmi and the sister line (color figure online)

Full size image

Inheritance of the resistance gene

To analyze the inheritance of resistance associated with Xa40(t) derived from the R-donor line, four populations were constructed by crossing cultivars Anmi and Ilpum as susceptible parents and 11,325 and 11,327 as resistant parents. The isolate HB01009 of Xoo race K3a was selected to inoculate the parents along with F1 progenies at the maximum tillering stage. The average lesion lengths of both susceptible parents (Anmi and Ilpum) were 15.45 ± 2.15 cm and 13.1 ± 1.53 cm at 14 days after inoculation (DAI), respectively (Supplementary Fig. 2). Resistant parents 11,325 and 11,327 were resistant to the K3a isolate, with an average lesion length of 0.51 ± 0.15 cm and 1.2 ± 0.35 cm at 14DAI. F1 plants developed from each combination showed moderate resistance to the K3a isolate, with an average lesion length of 4.5 ± 2.15 cm. The two mapping populations, Anmi/11,325 and Ilpum/11,325, were selected for genetic analysis of the gene since 11,325 showed a relatively stronger resistance. A total of 1583 F2 individuals from the two crosses were inoculated with K3a isolate at the maximum tillering stage under field conditions. The distribution of lesion lengths in the F2 populations was bimodal with an apparent valley at approximately 5 cm (Supplementary Fig. 3). Using this value (lesion length <5.0 cm) to differentiate between resistant and susceptible, the segregation ratio of F2 individuals from the cross Ilpum/11325 was 555 resistant and 241 susceptible, which is consistent with the expected allelic frequency of the 3:1 ratio (X 2 = 1.10, P > 0.05). Similarly, the segregation ratio of F2 individuals from the cross Anmi/11325 was 578 resistant and 209 susceptible, which was also consistent with the expected 3:1 ratio (X 2 = 1.03, P > 0.05, Table 2). This evidence indicates that the resistance of 11325 to BB race K3a is controlled by a single gene with incomplete dominance.

Table 2 Resistance response of F1 and F2 crosses between Ilpum and Anmi against race K3a
Full size table

Construction of a physical map spanning the Xa40 gene

A total of 54 molecular markers were selected and used to identify the markers linked to the target gene. A parental survey with R- and S-parents was carried out using PCR analysis. From the survey, 20 molecular markers were selected as anchor markers within the flanking region of RM1233 to RM5766 (Fig. 2a). A total of 80 F2 individuals showing a highly susceptible reaction (lesion length >12 cm) to the K3a isolate were selected from both cross combinations (Ilpum/11325 and Anmi/11325) to position Xa40(t) within the region. On the reference sequences of cv. Nipponbare released by IRGSP, 17 BAC clones were listed in the Nipponbare rice genome sequence, which covered the flanking region with a physical interval of 1.81 Mbp (Fig. 2a; Supplementary Table 1). The anchor markers landed on all except three of the BAC clones (OSJNBa0085H07, OSJNBb0049B20, and OSJNBa0004O15) with an average physical distance of about 86 kb between markers. Analysis of the number of distinct recombinants using genotypic data and phenotypic reaction showed that, among the 80 F2 individuals, nine and six distinct recombinants were identified within the region of interest. The marker RM27322, flanked by RM27320 and ID55.05-79, showed a co-segregation reaction in the population, allowing the position of the target gene to be narrowed down and mapped to a region within a 295-kb physical interval on chromosome 11.

Fig. 2
figure 2

Physical map of Xa40(t) conferring resistance to race K3a. a shows the 22 molecular markers located within the target region of 26.53–28.34 Mb on chromosome 11. Recombinant indicates a recombination event in the tested mapping population. The gray bar on the chromosome band corresponds to the physical region between 27 and 28 Mbp on chromosome 11. b is a fine map denoted by RM27320 and ID55.05-79 within a 295-kb physical interval. The numbers in parentheses indicate that the recombinants occurred at the corresponding marker loci derived in the resistant and susceptible plants. In the flanking region, a total 10 DNA markers are distributed. Based on these recombinants, Xa40 was delimited to an 80-kb interval between RM27320 and ID55.WA18-5, which is spanned by the BAC clone OSJNBa0036K13

Full size image

To further narrow the interval containing Xa40(t), we selected an additional 180 F2 individuals showing a susceptible reaction (lesion length ≥8 cm) from the same populations. In addition, one SSR marker, 20 STS markers, and one CAPS marker between RM27320 and ID55.05-79 were developed according to the sequence information from the rice genome database. Of these, markers displaying polymorphism between parents were chosen. The 10 markers, including the flanking markers (RM27320 and ID55.05-79), were used to detect the number of distinct recombinants in the newly selected 180 susceptible F2 individuals. Within the BAC clone OSJNBa0036K13, a total of one, zero, zero, and one recombinant from all the tested F2 individuals were identified with markers RM27320, RM27322, ID55WA3, and ID55.WA18-5, respectively. For OSJNBa0004O15, a total of two, two, and four recombinants were detected by markers LRR-1del, ID55.DEF48-5, and ID55.OsWa124, respectively. For OSJNBa0059H21, five recombinants were shown by ID55.05-79 (Fig. 2b). No recombinants were identified at markers RM27322 and ID55.WA3, which co-segregated with the Xa40(t) locus (Fig. 2b). Thus, based on the physical map of japonica rice Nipponbare, the Xa40(t) gene was defined by RM27320 and ID55.WA18-5 located on the BAC clone OSJNBa0036K13. The physical distance between the two markers is approximately 80 kb (Fig. 2b).

Candidate gene annotation

The intervening genomic region (~80 kb) of chromosome 11 where the Xa40 (t)gene was mapped contains at least eight gene models according to release 7 of the MSU Rice Genome Annotation Project Database and Resource (Table 3). The functionality of the genes predicted in the region includes one gene-encoding protein kinase, one encoding protein kinase domain-containing protein, one encoding expressed protein, one encoding wall-associated receptor kinase 3 precursor, three encoding transposon proteins, and one encoding retrotransposon protein. The region of Xa40(t) contains a gene (LOC_11g46900) encoding a wall-associated receptor kinase (WAK) 3 precursor. We considered WAK3, a promising candidate gene conferring resistance to BB race K3a.

Table 3 Annotation data of the putative ORFs in the target BAC clone (OSJNBa0036K13) region in the Nipponbare genome database
Full size table

Semi-quantitative RT-PCR analysis

To compare the expression level of the eight genes identified in the target region, a primer set (ID55.WA3) based on coding DNA sequence (CDS) was designed (Supplementary Table 2). The semi-quantitative PCR analysis of resistance to K3a race-related genes was performed using cDNA synthesized from the leaf of each sample collected at 0, 1, 4, 8, 24, and 72 h after BB inoculation. Expression patterns of the genes were observed in Anmi and 11325, respectively (Fig. 3). The accumulation of WAK3 messenger RNA (LOC_11Os11g46900) was strongly promoted by inoculation in leaves of resistant 11325, but was not induced in the susceptible cultivar Anmi. The transcript levels of LOC_11g46880 were induced only in the susceptible line. Apart from these two genes, there were no significant differences between the resistant and susceptible lines in the transcript levels of the genes tested. The mRNA induction of LOC_11Os11g46900 was also pronounced during the procedure of leaf sampling for RNA extraction, indicating that the WAK3 gene may be induced in response to wound stress by leaf-clipping.

Fig. 3
figure 3

Semi-quantitative RT-PCR assessment of gene transcript levels in the leaf tissue from resistant (11325) and susceptible (Anmi) genotypes following 0–48 h of inoculation. Rice actin1 was used as an internal control in the same PCR condition

Full size image

Haplotype test

In the haplotype test using the STS primer ID55.WA3, only seven lines out of 24 cultivars/lines showed the same size of allele, A1, against 11325 with Xa40(t) (Table 4). The PCR products of the A1 type were recovered for sequencing by cloning and a single nucleotide polymorphism was found in A1 alleles, allowing classification into the G and A type. Among the four ancestors of IR65482-7-216-1-2, none showed 11325-type allele A1_A (Table 4). This SNP was confirmed by enzyme digestion using HpCH4III on 1.2 % agarose gel (Supplementary Fig. 4). Tracking the source of this gene could not be undertaken in this study as we did not have access to seeds of all of the ancestors in the pedigree.

Table 4 Haplotypes of the Xa40(t) locus based on alleles of the WAK3 gene in rice cultivars
Full size table

Discussion

BB disease caused by Xoo is one of the major constraints for rice (O. sativa L.) production in most rice-growing areas worldwide. The disease is vascular, resulting in a systemic infection that produces gray to white lesions along the veins (Nelson et al. 1994). The most effective control method considered for rice BB disease is to use resistant cultivars. To date, around 39 BB resistance genes have been mapped on different rice chromosomes, and nine R-genes have been isolated using map-based cloning (Dai et al. 2007; Wang et al. 2014a). Many resistance genes have broken down following their extended use over wide areas under conditions conducive to disease development and intense selection pressure for virulent races of the pathogen. Therefore, new approaches using R-gene pyramiding, or a novel resistance gene with multiple-resistance specificity from a single locus, are necessary to provide stronger and more diversified levels of resistance as well as sustainable durable resistance (Gu et al. 2005; Chen et al. 2011). Here, we report evidence of the identification of a new BB gene conferring resistance to four BB isolates endemic to Korea.

Resistant lines 11325 and 11327 used in this study were developed from a backcross breeding program for the introgression of the Bph18 gene to a japonica background using IR65482-7-261-2 derived from O. australiensis (Suh et al. 2011). To date, there are no known BB resistance genes in the two lines or the donor, even though the lines proved resistant to all races: K1, K2, K3, and K3a (Table 1). The R-gene Xa3 is the main R-gene deployed against BB in resistant cultivars throughout Korea. The occurrence of a new BB pathotype underscores the need for new resistance genes.

Analysis of the inheritance of Xa40(t) in F1 plants from three populations of Shindongjin/11325(data not shown) showed that the F1 plants exhibited an intermediate resistance—more resistant than susceptible, but less than the resistant parent (Table 2)—and thus the resistance was considered to be inherited by a gene with incomplete dominance. This inheritance pattern is similar to the case of the xa33(t) gene in rice (Korinsak et al. 2009). It is frequently reported in rice resistance studies that phenotype was controlled by a dominant or recessive gene with incomplete dominance (Ise et al. 2002; Zhang et al. 2006). The segregation ratio of F2 individuals from both populations was shown to fit the expected 3(R):1(S) ratio when the resistance level was determined by lesion length <5.0 cm (Table 2). The R-gene designated as Xa40(t) was flanked by RM1233 and RM5766, covering around 27.0–28.9 Mbp of the physical region on the long arm of chromosome 11, along with 17 rice BAC clones (Fig. 2; Supplementary Table 1). According to previous reports, six BB R-genes (Xa21, Xa4, Xa3/Xa26, Xa22, Xa10, and Xa23) were already reported on chromosome 11L; in particular, Xa4 and Xa3/Xa26 were clustered together and located close to each other in that region (Song et al. 1995; Iyer and McCouch 2004; Sun et al. 2004; Verdier et al. 2012; Wang et al. 2014a). To discriminate these genes from Xa40(t), a validation test was performed with DNA markers specifically able to detect the genes. In this PCR analysis, none of the lines showed amplicons relating to R-gene alleles of Xa3/Xa26 and Xa4 (data not shown). In phenotypic response to BB, the susceptible japonica parents Anmi, Shindongjin, and Junam, which include BB resistance gene Xa3, all showed a susceptible reaction to K3a isolate (Table 1) and the Xa3 indica monogenic line IRBB3 also exhibited a susceptibility to K3a (Table 4). Both IRBB4 and japonica NILs, including the Xa4 gene, showed moderate resistance to the isolate. Based on these observations, we decided to conduct an allelism test to confirm the difference between Xa4 and Xa40(t), even though the two genes had different responses to the K3a isolate, showing moderate resistance and resistance, respectively, in this study. A total of 867 F2 plants from the cross between IRBB4 (with Xa4) and 11325 were inoculated with K3a isolate and the segregation pattern of resistant (lesion length below 5 cm) and susceptible plants had a ratio of 13R:3S (X 2 = 5.76, P > 0.05). Physically, both genes are closely linked with each other (approximately 0.5 Mbp). Given this fact, it is difficult to expect this many susceptible lines by chromosomal crossover. From the result, it is possible to assume that the occurrence of susceptible lines of high frequency resulted from a gene peculiar to Xa4 depending on zygosity type. The response of the Xa4 gene to K3a varied depending on allele type and gene combination (Suh et al. 2013). In particular, it has shown an unstable reaction to the race in heterozygous conditions.

Even though the allelism test carried out before fine-mapping did not give a clear answer because of the complexity caused by gene recombination, some evidence such as the difference in phenotypic reaction to the same race, a validation test using gene-specific markers (Supplementary Table 3), and haplotype analysis indicate that the R-gene in 11325 is a different allele from Xa4.

In this study, the region was narrowed down to the BAC clone regions flanked by markers RM1233 and RM5766 on chromosome 11L. Using recombinant analysis, we defined an interval of approximately 295 kb flanked by the ORF-based markers (Supplementary Table 2). Through additional analysis of recombination events within 180 F2 plants, the Xa40(t) gene was further narrowed down to an 80-kb region on chromosome 11L. Sequence data from the region harboring the Xa40(t) gene indicated that this region contained eight candidate genes (Table 3), which all differed from other Xoo resistance genes reported to date. Given that the products of the other four BB R-genes that have been cloned are unique and not found in other plant species (Xa21 and Xa26 encode for similar receptor-like proteins; Dai et al. 2007), this result suggests a new type of BB resistance gene. In semi-quantitative PCR analysis confirming the gene associated with BB resistance, the transcript level of LOC_Os11g46900 in 11325 displayed a tendency to gradually increase over time after inoculation; whereas, this gene was not induced in the susceptible cultivar Anmi (Fig. 3). With reference to the mRNA induction of the gene at 0 h after inoculation, the patterns by mock inoculation were identical with each other regardless of the time after inoculation (data not shown). The wall-associated kinases, or WAKs, are receptor-like kinases that are linked to the pectin fraction of the cell wall and have a cytoplasmic protein kinase domain (Kohorn and Kohorn 2012). This ability to bind and respond to several types of pectin correlates with a demonstrated role for WAKs in both pathogen response and cell expansion during plant development. WAK expression is induced by wounding, pathogen infection, and by many stresses such as ozone and heavy metals (He et al. 1998; Anderson et al. 2001; Kohorn 2001; Wagner and Kohorn 2001; Sivaguru et al. 2003). We are not sure whether the WAK3 gene confers resistance to BB in this study. Despite this, we consider that this study suggests a potential new BB R-gene, and results reported in this study, coupled with existing knowledge, provide evidence for this case. In a further study, we plan to identify the function of candidate genes through advanced methods, synthesis of the full-length cDNA of this gene and gene silencing.

In Korea, the resistance of the Xa3 gene prevailing in most resistant Korean cultivars was broken down by the occurrence of a new BB race, K3a. This paper reports a new source of resistance to race K3a, the subsequent fine-mapping of the resistance gene to a region on rice chromosome 11, and the identification of a candidate gene underpinning resistance. The identification of Xa40(t) conferring resistance to race K3a is of significant value to rice breeding programs through the expansion of the pool of R-genes conferring resistance to local races of Xoo and the use of advanced backcross lines with the gene for marker-assisted breeding and gene pyramiding strategies. Further characterization of Xa40(t) at the protein level will be helpful to elucidate the mechanisms of resistance.

Author contribution statement

SMK (s.kim@irri.org) carried out the development of the mapping population, genetic analysis, data gene mining, and primer design, and drafted the manuscript. JPS (suhjp@rda.go.kr) participated in the phenotypic selection and development of the backcrossed lines. YQ (qinyang2013@korea.kr) performed cDNA synthesis and semi-quantitative reverse transcription PCR. THN (nohtw831@rda.go.kr) contributed to the isolation and culture of virulent strains of Xoo used in this study. RFR (r.reinke@irri.org) was involved in phenotypic selection for fine-mapping and data mining of candidates. KKJ (k.jena@irri.org) designed and coordinated the study, assisted with genetic analyses, and drafted the manuscript. All authors read and approved the final manuscript.