Introduction

Brown planthopper (BPH), Nilaparvata lugens Stål (Homoptera: Delphacidae), is one of the most destructive insect pests of rice (Oryza sativa L.), especially of temperate japonica rice cultivars where few resistance genes against BPH have been incorporated (Jena et al. 2006). BPH extracts phloem saps of rice plants using its stylet-type mouthparts, resulting in a severe damage symptom known as ‘hopper-burn’. BPH also transmits rice grassy stunt virus and ragged stunt virus as vectors (Rivera et al. 1966; Heinrichs 1979). On the basis of differential reactions, BPH populations in different countries have been categorized into four biotypes (Khush et al. 1985). The original populations in East and Southeast Asia belonged to biotype 1. Biotype 2 originated in Phillippines, Indonesia and Vietnam (Khush 1979) and is the dominant biotype in these countries. Biotype 3 was produced in the laboratory at the International Rice Research Institute (IRRI) (Pathak and Khush 1979) and in Japan (Ikeda and Vaughan 1991). Biotype 4 is found only in South Asia (Khush 1984). In Southeast Asia, the BPH populations shifted from biotype 1 to biotype 2 in the 1970s (Feuer 1976; Mochida et al. 1977; Stapley et al. 1979), and at present comprise mostly a complex of biotypes 2 and 3 (Medrano and Heinrichs 1985; Sogawa et al. 1987; Huynh and Nhung 1988). On the other hand, Chelliah and Bharathi (1993) categorized BPH populations into five biotypes on the basis of their differential reactions to a set of reference cultivars. Of the biotypes (1, 2, and 3) known to be present in Korea, biotype 1 is the most prevalent (Park and Song 1988). Conventional methods of controlling BPH are highly dependent on spraying poisonous chemicals, which is expensive in terms of labor, money, and the environment. In addition, the occurrence of resurgence, a phenomenon of pest population increase after application of insecticides (Heinrichs et al. 1982), is problematic. Thus introduction of resistant cultivars is beneficial economically and environmentally (Huang et al. 2001).

Incorporating resistance gene(s) from wild species into cultivated species can be an alternative approach to develop BPH resistance in susceptible commercial cultivars. In general, wild species are poor in important agronomic traits such as yield, plant type, grain type, eating quality, seed shattering habit, etc. However, the wild species of Oryza are regarded as a treasure trove of novel genes for resistance to disease or insect pests and tolerance against environmental stresses. Since few useful genes from wild germplasm accessions have been explored, there is still great potential for exploring novel genes. Several wild species including O. minuta, O. latifolia, O. nivara, O. officinalis and O. punctata were reported to possess resistance genes to different BPH biotypes (Wu et al. 1986). Nine resistance loci among a total of 19 BPH resistance loci reported so far in rice have been identified from wild species. They are Bph10 on the long arm of chromosome 12 from O. australiensis (Ishii et al. 1994), Bph12(t) on the short arm of chromosome 4 from O. latifolia (Yang et al. 2002), Bph13(t) on the long arm of chromosome 2 from O. eichingeri (Liu et al. 2001), another Bph13(t) against BPH biotype 4 on the short arm of chromosome 3 from O. officinalis (Renganayaki et al. 2002), Qbp1 and Qbp2 (later named as Bph14 and Bph15, respectively) on the long arm of chromosome 3 and the short arm of chromosome 4, respectively, from O. officinalis (Huang et al. 2001), Bph18(t) on the long arm of chromosome 12 from O. australiensis (Jena et al. 2006), and bph11(t) and bph12(t) on the long arm of chromosome 3 and chromosome 4 respectively, from O. officinalis (Hirabayashi et al. 1999).

O. minuta (2n = 48, BBCC genome, Acc No.101141), an allotetraploid wild species, belonging to the O. officinalis complex, is endemic to Philippines and Papua New Guinea. It is known to have useful genes for resistance to bacterial blight (BLB), rice blast, and BPH (Brar and Khush 1997). However, few studies have been conducted to transfer the resistance genes from O. minuta to commercial rice cultivars. Amante et al. (1998) evaluated the resistance of advanced progenies introgressed from O. minuta for resistance against BLB and rice blast disease. The introgression line IR71033-121-15 derived from the cross of O. sativa/O. minuta showed resistance to BPH biotypes of Korea (Jena KK, unpublished). However, molecular mapping of these resistance gene(s) has not been reported.

The present study was conducted to identify the BPH resistance genes in IR71033-121-15 and to develop the markers for use in breeding BPH resistance in rice.

Materials and methods

Plant materials

The introgression line, IR71033-121-15, derived from an interspecific cross between IR31917-45-3-2 and a wild species O. minuta (Acc. No. 101141) showed strong resistance to the Korean BPH biotype 1. IR71033-121-15 and Junambyeo (a Korean japonica cultivar susceptible to BPH) were used as parents to develop an F2 population of 190 plants for genetic analysis. The 190 F3 lines harvested from each of the F2 plants were bioassayed for BPH resistance. IR31917-45-3-2 (progenitor of IR71033-121-15) and two Tongil-type rice cultivars, Taebakbyeo and Andabyeo, were used as check cultivars in the bioassays. IR31917-45-3-2, Taebakbyeo, and Junambyeo are susceptible to Korean BPH whereas IR71033-121-15 and Andabyeo are resistant. Seeds of IR71033-121-15, IR31917-45-3-2, and O. minuta were obtained from the Plant Breeding and Genetics and Biotechnology Division of IRRI, Los Baños, Philippines. Seeds of Junambyeo, Taebakbyeo and Andabyeo were obtained from Rice Research Division of National Institute of Crop Science (NICS), Rural Development Administration (RDA), Korea.

Bioassay for BPH resistance

A pure BPH population developed from Korean BPH biotype 1 was used for bioassay of parents and F2/F3 population. Bioassays were conducted at the greenhouse facility of the Rice Research Division of NICS, RDA. The bioassay was performed with a modified bulk seedling test following the method of Pathak et al. (1969). Fifteen seedlings from each of 190 F3 lines were planted in a row per line with three replications. Junambyeo and Taebakbyeo were used as susceptible (S) checks, and IR71033-121-15-2 and Andabyeo were used as resistant (R) checks. Seedlings at the two and a half leaf stage were infested with second or third-instar nymphs at a density of 10–12 nymphs per seedling. The reaction against the BPH was scored following the guidelines of Standard Evaluation Systems for Rice (IRRI, 1988): 0, no damage; 1, very slight damage; 3, first and second leaves of most plants partially yellowing; 5, pronounced yellowing and stunting or about 10 to 25% of the plants wilting; 7, more than half of the plants wilting or dead and remaining plants severely stunted or dying; 9, all plants dead.

PCR amplification and marker detection

Plant DNA was extracted from the frozen leaves of rice plants using the CTAB method (Murray and Thompson 1980). For PCR amplification of markers, each 20 μL reaction mixture contained 50 ng DNA, 5 pmol of each primer, 10× PCR buffer [100 mM Tris (pH 8.3), 500 mM KCl, 15 mM MgCl2, 2 μg gelatin], 250 μM of each dNTP and 0.5 U of Taq polymerase. PCR was performed in a PTC-100 Programmable Thermal Controller thermocycler (MJ Research Inc, Waltham, MA). The thermocycler profile was: 5 min at 94°C, 35 cycles of 1 min at 94°C, 1 min at 48°C (for STS) or 55°C (for SSR), and 2 min at 72°C, with a final extension of 5 min at 72°C. Amplified PCR products were resolved by electrophoresis in 3% agarose gels with ethidium bromide staining or 8% polyacrylamide denaturing gels with silver-staining for SSR markers (Panaud et al. 1996).

Construction of a framework map, detection of QTLs for BPH resistance and high-resolution mapping

Linkage analysis was conducted using MAPMAKER version 3.0 software (Lander et al. 1987). Map distances were estimated by the Kosambi function (Kosambi 1944). The linkage map in this study was basically the same as previously reported (Rahman et al. 2007). In this study, 46 more F2 plants were added to construct a framework map. The M-QTL and E-QTLs between random marker intervals were determined using QTLMapper version 2.0 with the mixed linear model approach (Wang et al. 1999). After detecting QTL positions for BPH resistance based on LOD and R2 values, fine mapping of the resistance loci was performed with SSR and STS markers. Graphical data of the recombinants between markers flanking the target regions were analyzed to clarify QTL positions, and introgressed segments from O. minuta were checked at the QTL regions for confirmation of the QTLs. SSR markers located within the two flanking markers of the resistance QTL regions were adopted from the public database released by the International Rice Microsatellite Initiative (IRMI, http://www.gramene.org/microsat), and PCR-based STS markers were developed according to the sequence information which is available at http://www.ncbi.nlm.nih.gov/ (for indica) and at http://www.rgp.dna.affrc.go.jp/ (for japonica) as described in Chin et al. (2007). For developing new SSR markers, the putative simple sequence repeat motifs were searched and identified using the Simple Sequence Repeat Identification Tool (SSRIT, http://www.gramene.org/db/searches/ssrtool) and for new STS markers, the putative sequences which were highly divergent between the two reference genomes were targeted. All PCR primers were designed using Primer3 version 4.0 (http://frodo.wi.mit.edu/). The markers used for fine mapping of the target regions are listed in Table 1.

Table 1 Summary of PCR-based markers developed for two specific QTLs
Full size table

Physical map construction

The physical map of the target QTLs was constructed by bioinformatics analysis using bacterial artificial chromosome (BAC) and P1-derived artificial chromosome (PAC) clones of cv. Nipponbare released by the International Rice Genome Sequencing project (IRGSP, http://rgp.dna.affrc.go.jp/IRGSP/index.html). The clones were anchored with the target gene-linked markers and then alignment of sequences was carried out using pairwise BLAST (http://www.ncbi.nlm.nih.gov/blast/bl2seq/b12.html).

Results

Inheritance of BPH resistance in “IR71033-121-15”

The introgression line IR71033-121-15 expressed strong resistance to the Korean BPH biotype 1, while the recurrent parent, Junambyeo, was completely susceptible to BPH (Fig. 1). The F1 plants from the cross between Junambyeo and IR71033-121-15 showed partial resistance to BPH, indicating that BPH resistance of IR71033-121-15 was controlled by more than one dominant gene and the dominance was not complete. The resistance scores of the 190 F3 lines infested with BPH ranged from 0.0 to 9.0 and showed a continuous variation with two peaks around 1.5 and 3.5 in the segregation curve (Fig. 1), demonstrating the involvement of major genes or QTLs controlling the segregation of resistance.

Fig. 1
figure 1

Frequency distribution of the brown planthopper (BPH) resistance scores among F3 families from the cross Junambyeo/IR71033-121-15

Full size image

Linkage map construction and identification of QTLs for BPH resistance

A linkage map was constructed by genotyping 190 F2 individuals using 143 markers. Using the QTLMapper version 2.0, two chromosomal loci for the resistance were identified on the short arm of chromosome 4 delimited by MS10 and RM5953 (3.5 cM distance) and the long arm of chromosome 12, delimited by RM3726 and RM5479 (4.1 cM distance) (Table 2; Figs. 2, 3). At both loci, alleles from the resistant parent IR71033-121-15 significantly increased BPH resistance. These two QTLs were tentatively designated as QTL-4 on chromosome 4 and QTL-12 on chromosome 12. The R 2 values of QTL-4 and QTL-12 were fairly high representing 26.6% and 14.5% of phenotypic variation, respectively, and the summed R 2 of the two QTLs was 41.1% (Table 2). In addition, a total of two significant epistatic interactions between random marker intervals were identified which included a significant interaction between the two QTLs (Table 2). Total phenotypic variation explained by these QTLs and significant interactions was 47.2%. However, there was no dominance effect in the two major QTLs and other marker loci, explaining the intermediate resistance phenotype of F1 plants shown in Fig. 1. The allele effect of each major QTL on BPH resistance was greater in QTL-4 than in QTL-12.

Table 2 Major QTLs and significant digenic epistatic interactions identified for BPH resistance in F2:3 from a cross between Junambyeo and IR71033-121-15
Full size table
Fig. 2
figure 2

a Genetic and physical map of the Bph20(t) locus on the short arm of chromosome 4. Positions of QTLs (blue vertical bars) previously reported near the Bph20(t) locus were estimated based on the sequence analysis of markers flanking the QTLs using Gramene DB (http://www.gramene.org/): Bph12 from O. latifolia (Yang et al. 2002); Bph15 from O. officinalis (Huang et al. 2001; Yang et al. 2004); Bph17 from Rathu Heenati (Sun et al. 2005). b Phenotype (BPH resistance) and graphical genotype of selected F 3 lines showing crossovers near by the Bph20(t) locus. I IR71033 allele, J Junambyeo allele, H heterozygous

Full size image
Fig. 3
figure 3

a Genetic and physical map of the Bph21(t) locus on the long arm of chromosome 12. Positions of QTLs (blue vertical bars) previously reported near the Bph21(t) locus were estimated based on the sequence analysis of markers flanking the QTLs using Gramene DB: Bph10 from O. australiensis (Ishii et al. 1994), Bph18(t) from O. australiensis (Jena et al. 2006). b Phenotype (BPH resistance) and graphical genotype of selected F3 lines showing crossovers near by the Bph21(t) locus. I IR71033 allele, J Junambyeo allele, H heterozygous

Full size image

Position of QTLs on the physical map

All the anchor markers used for fine mapping of the two resistance loci were landed on the reference sequences of cv. Nipponbare by bioinformatics analysis using a software tool BLASTN. The 1.31 Mb virtual contig map of chromosome 4 composed of 14 overlapping BAC/PAC clones was constructed to locate the physical regions for markers MS10 and RM5953 and thus delimit the QTL-4 locus. Similarly, the 1.13 Mb virtual contig map of chromosome 12 composed of 13 overlapping BAC/PAC clones was constructed to locate the physical regions for markers RM3726 and RM5479 and delimit the QTL-12 locus. Additional markers were developed using publicly available sequence databases aided by bioinformatics analysis to narrow down the physical distance between the two flanking markers (Table 1). Finally, two saturated genetic maps and subsequent contig maps were constructed for QTL-4 on the short arm of chromosome 4 (Fig. 2) and QTL-12 on the long arm of chromosome 12 (Fig. 3).

For QTL-4, the markers B43, S4019A, and S4019B derived from OSJBa0028M15, flanked by B42 and B44, co-segregated with the BPH resistance phenotypes. Association of marker genotypes with phenotypes (BPH resistance) of F3 lines that contained the crossover event on the QTL-4 region was analyzed (Fig. 2b). Based on comparisons of the genotypes at both QTL-4 and QTL-12 together with the phenotypes, the lines 73 and 91 delimited the left margin of the QTL-4 position, and similarly, the lines 76 and 115 delimited the right margin of the QTL-4 position. From these results, it was concluded that QTL-4 resides between B42 and B44, which spans 1.2 cM and corresponds to a 193.4-kb region in the reference physical map (http://www.rgp.dna.affrc.go.jp/) (Fig. 2a).

For QTL-12, the markers B120, S12094B and B121 derived from OJ1310_CO3, flanked by S12094A and B122, co-segregated with BPH resistance. Using the same procedure as for QTL-4, the association of marker genotypes with BPH resistance of F3 lines containing the crossover event in the QTL-12 region was analyzed (Fig. 3b). The lines 13 and 28 delimited the left margin of the QTL-12 position, and the lines 43 and 99 delimited the right margin of the QTL-12 position. Therefore, it was concluded that QTL-12 resides between 12094A and B122, which spans 1.0 cM and corresponds to a 194.0 kb region in the reference physical map (Fig. 3a).

Evidence of O. minuta chromosome segment integration in two putative QTLs

To confirm that the two chromosomal regions containing resistance QTLs were introgressed from the wild species O. minuta, the alleles of markers B43, S4019A, S4019B, and B44 on chromosome 4, and B120, S12094B, B121, and B122 on chromosome 12 were compared between IR71033-121-15 and O. minuta. In the QTL-4 region, the same PCR amplicon (≥200 bp) was produced for markers S4019A and S4019B in both O. minuta and IR71033, which was different from the amplicon generated in the recurrent parent IR31917-45-3-2 and Junambyeo (Fig. 4a). However, B43 at the upper position and B44 at the lower position were negative for introgression. Similarly, analysis of markers in the QTL-12 region revealed that the same PCR amplicons were produced for markers S12094B and B121 in IR71033-121-15 and O. minuta, which were different from the amplicons generated in the recurrent parent IR31917-45-3-2 and Junambyeo (Fig. 4b). However, B120 at the upper position and B122 at the lower position were not positive for introgression. These results indicate that the two resistance QTLs correspond to the introgression of O. minuta DNA into chromosome 4 (S4019A to S4019B) and 12 (S12094B to B121) of IR31917-45-3-2.

Fig. 4
figure 4

Introgression test of O. minuta segments into IR71033-121-15 a at the QTL4 locus on chromosome 4, and b at the QTL-12 locus on chromosome 12. S Size marker, I IR71033-121-15, M O. minuta (accession # 101141), J Junambyeo, R IR31917-45-3-2. An arrow indicates the B120 marker

Full size image

Discussion

Nineteen major BPH resistance loci have been reported in indica cultivars along with four wild species, O. australiensis, O. eichingeri, O. latifolia, and O. officinalis. Of these, 17 resistance loci (Bph1, bph2, Bph3, bph4, Bph6, Bph9, Bph10, bph11, bph12(t), Bph12, Bph13(t), Bph13, Bph14, Bph15, Bph17, Bph18, and bph19) have been assigned to rice chromosomes (Ikeda 1985; Ishii et al. 1994; Hirabayashi and Ogawa 1995; Hirabayashi et al. 1999; Murata et al. 1998, 2001; Kawaguchi et al. 2001; Liu et al. 2001; Jena et al. 2002, 2006; Renganayaki et al. 2002; Sharma et al. 2003, 2004; Yang et al. 2004; Kim and Sohn 2005; Sun et al. 2005; Chen et al. 2006; Jairin et al. 2007). QTL studies involving the BPH-resistant cultivars IR64, Kasalath, DV85, Teqing, Col.5, and B5 introgression line created through the introgression of wild rice O. officinalis have also been carried out (Alam and Cohen 1998; Huang et al. 2001; Su et al. 2002, 2005; Xu et al. 2002; Ren et al. 2004; Soundararajan et al. 2004). Candidate genes for BPH resistance have been reported for the indica cv. Samgangbyeo (Park et al. 2007).

The main goal of this study was to identify new BPH resistance loci originating from the wild species, O. minuta, which has the BBCC genomes. Our analysis resulted in the identification of two major BPH resistance QTLs on chromosomes 4 and 12. These QTLs together explained over 40% of the observed phenotypic variance. Studies involving BPH resistance introgressed from other wild species have also resulted in the identification of loci mapping to chromosome 4. Previously, Huang et al. (2001) reported that two QTLs for BPH resistance, introgressed from O. officinalis (CC), were located on the short arm of chromosome 4 and long arm of chromosome 3. Later, Yang et al. (2004) identified one BPH resistance locus on chromosome 4 (Bph15) using the same introgressed line from O. officinalis. Yang et al. (2002) also reported that a BPH-resistant gene, Bph12(t) from O. latifolia (CCDD), was located on the short arm of chromosome 4. Of the Bph genes that have been mapped to chromosome 4, Bph17 from the Sri Lankan cultivar Rathu Heenati is the closest to QTL-4 described in our study (Fig. 2a). Bph17 is reportedly located between two SSR markers RM8213 and RM5953 with map distances of 3.6 cM and 3.2 cM, respectively (Sun et al. 2005). Part of this region covers the QTL-4 locus; however, the fact that QTL-4 originates from the wild species O. minuta suggests Bph17 is a different gene. Further studies such as the cloning of these genes will clarify their relationship.

For the long arm of chromosome 12, six resistance loci have been reported: Bph1 from ‘Mudgo’, ‘MTU15’, ‘Co22’, ‘MGL2’, ‘Samgangbyeo’ (Kim and Sohn 2005) and ‘Gayabyeo’ (Jeon et al. 1999); bph2 from ‘Karsamba Red, ASD7’ (Murata et al. 1998); Bph9 from ‘Pokkali’ (Murata et al. 2001) and Kaharamana (Su et al. 2006); Bph10 (Ishii et al. 1994),and Bph18(t) (Jena et al. 2006) from O. australiensis. Of these loci, Bph10 and Bph18(t) are located nearby QTL-12. However, as with QTL-4, comparative analysis of marker loci revealed that these resistance genes actually map apart from QTL-12 (Fig. 3a). Given that the two major QTLs in this study differ in chromosomal location and/or origin (O. minuta) from BPH resistance loci reported previously, we have designated these loci as Bph20(t) and Bph21(t) on chromosome 4 and 12, respectively. This is the first report on BPH resistance loci introduced from a BBCC genome wild species.

It is notable that BPH resistance genes are clustered in the same regions of chromosome 4 and 12 despite their different origins. Similar findings were also observed in the case of BPH resistance loci Bph13(t) (Renganayaki et al. 2002) and the recessive bph19(t) clustered on the short arm of chromosome 3 (Chen et al. 2006), and Bph10 (Ishii et al. 1994) with Bph18(t) (Jena et al. 2006) on the long arm of chromosome 12.

Besides the two major QTLs, significant epistatic interactions between two random markers were found to play a certain role in the expression of resistance to BPH (Table 2). This suggests that epistatic interactions as well as major QTLs should be taken into consideration for breeding BPH-resistant cultivars through MAS, although this might complicate the MAS process (Qiao et al. 2008). The existence of an additive effect for the two major QTLs reported here indicates that lines with stronger BPH resistance can be developed by incorporating multiple QTLs. We are now developing resistant japonica lines by backcrossing the resistant progenies with the recurrent parent Junambyeo using MAS for Bph20(t) and Bph21(t).

Of the five biotypes reported (Chelliah and Bharathi 1993), resistance genes have been identified against four of them (Panda and Khush 1995). Bph1 confers resistance against biotypes 1 and 3, bph2 is closely linked with Bph1 and provides resistance against biotypes 1 and 2, Bph3 and bph4, which are closely linked, confer resistance against four biotypes, and Bph5, Bph6, and Bph7 provide resistance against only biotype 4. The relationship between other resistance genes and corresponding BPH biotypes has not been comprehensively documented. Considering the potential changes of BPH biotypes, due to the extensive cultivation of high yielding rice cultivars and different environmental factors, new resistance genes need to be identified to ensure the durability of host resistance. Wild species of rice are important genetic resources for breeding insect resistant cultivars. The parental line, IR71033-121-15, used in this study carries two new resistance genes, Bph20(t) and Bph21(t), introgressed from O. minuta, and is expected to be a valuable source of BPH resistance.