Evaluation and breeding application of six brown planthopper resistance genes in rice maintainer line Jin 23B

Background The brown planthopper (BPH), an insect species that feeds on rice plants (Oryza sativa L.), is one of the most destructive insect pests of rice. BPH can be controlled with chemical pesticides, but they are expensive in terms of the cost and environmental hazards. Hence, utilization of resistance genes in resistant varieties is considered as a more economical and eco-friendly effective method for control. Results In this study, six dominant BPH-resistance genes (Bph3, Bph14, Bph15, Bph18, Bph20 and Bph21) were introduced into an elite indica rice cultivar Jin 23B using the marker-assisted selection (MAS) method of breeding. One line combining Bph14 and Bph15 and six single gene introgression lines were used to evaluate the gene effects based on three parameters: seedling and tillering resistance of the rice genotypes, honeydew weight, and BPH survival rate. Among all improved lines, combination of Bph14 and Bph15 had the largest effect in conferring resistance to BPH. Bioassays showed that the order of the gene effects was Bph14/Bph15 > Bph15 ≥ Bph14 ≥ Bph20 ≥ Bph21 ≥ Bph3 > Bph18 > none at the rice seedling stage. The pyramided or single-gene introgression lines showed enhanced resistance relative to the control. Furthermore, field trial data demonstrated that yields of improved Jin 23B lines were similar to the control under BPH-free field conditions. Conclusions Each of the BPH resistance genes reduced BPH growth and development, and was effective at both the seedling and tillering growth stages. These lines can be used in rice hybrid and production in China.


Background
Brown planthopper (Nilaparvata lugens Stål, BPH), is one of the most destructive insect pests of tropical and temperate rice (O. sativa L.) in Asia. BPH extracts phloem saps from host plants using its stylet-type mouthparts, resulting in severe and damaging symptom known as 'hopper-burn' (Watanabe and Kitagawa 2000). BPH also transmits viruses (Cha et al. 2008), such as grassy stunt virus and rugged stunt virus to rice plants, causing further yield losses in Asian countries, including China, Japan, Korea and Vietnam (Brar et al. 2010). In rice production, BPH is managed mainly by using insecticides that are expensive in terms of cost and environmental hazards. Therefore, utilization of BPH resistance genes in developing resistant varieties is considered as the most economical and eco-friendly approach to control this pest (Tanaka et al. 2000).
More than 32 BPH resistance loci have been identified in indica rice cultivars and wild rice species (Ali and Chowdhury 2014;Wang et al. 2015). Several of these resistance genes have been assigned to rice chromosomes 2, 3, 4, 6, and 12 (Jena and Kim 2010) and at least ten genes have been fine-mapped to genomic region of less than 200 kb (Hu et al. 2016). Most of the resistance alleles are dominant, but a few are documented as recessive (bph4, bph5, bph7, bph8, bph19,bph25 and bph29). Seven BPH-resistance genes Bph14, Bph26, Bph17, bph29, Bph9, Bph18 and Bph32 have been cloned by map-based cloning (Du et al. 2009;Tamura et al. 2014;Liu et al. 2015;Wang et al. 2015;Zhao et al. 2016;Ji et al. 2016;Ren et al. 2016). Among the resistance loci, the broad-spectrum resistance gene Bph3, identified in cultivars Rathu Heenati and PTB33, was mapped on the short arm of chromosome 6 where it was flanked by SSR markers RM589 and RM588 (Jairin et al. 2007a). Bph14 and Bph15 were previously mapped in introgression line B5 (derived from O. officinalis), with QBph1 (renamed as Bph14) on the long arm of chromosome 3 explaining 26. 4% of the variation in BPH response and Qbph2 (renamed as Bph15) on the short arm of chromosome 4 explaining 14.3% of the variation . Bph14, encoding a coiled-coil, nucleotide-binding, and leucine-rich repeat protein was cloned using a map-based cloning approach (Du et al. 2009). Bph15 was fine mapped to an interval of approximately 50 kb between map positions 6,883,264 and 6,934,420 (Yang et al. 2004). The Bph18 gene, originally derived from O. australiensis, was mapped in indica introgression line IR65482-7-216-1-2, shown to be on the long arm of chromosome 12, and explained 47% of the variation in BPH response (Jena et al. 2006). Mapbased cloning and complementation test revealed that Bph18 encodes a CC-NBS-NBS-LRR protein and is a functionally different allele of Bph26 (Ji et al. 2016). Bph20 and Bph21 in introgression line IR71033-121-15 were derived from O. minuta (Accession number 101141), with Bph20 located to a 193.4 kb region on the short arm of chromosome 4, and with Bph21 located to a 194 kb region on the long arm of chromosome 12 (Rahman et al. 2009).
Near-isogenic lines (NILs) and gene combination lines carrying one or several resistance genes have been developed (Sharma et al. 2004;Hu et al. 2012Hu et al. , 2013 and in an effective manner to facilitate the use of the resistance genes or of quantitative trait loci (QTLs) for future rice improvement program. Bph1, Bph3 and Bph4 have already been successfully introduced into some elite cultivars using MAS and backcross breeding programs at the International Rice Research Institute (IRRI) in the Philippines (Khush and Brar 1991;Jena and Kim 2010). Qiu et al. (2011) used MAS to pyramid Bph6 and Bph12 into a japonica line and an indica line. Results of BPH bioassays indicated that the pyramided line showed higher resistance than the Bph6or Bph12-single introgression line. Hu et al. (2012) incorporated Bph14 and Bph15 genes into Minghui 63 and its hybrid Shanyou 63, and showed that hybrids containing a single BPH resistance gene had enhanced resistance compared to conventional hybrids that lack of resistance gene. Lines combining two genes were more resistant than the respective single gene lines. Hu et al. (2013) incorporated Bph14, Bph15 and Bph18 into cv. 9311 and its hybrids GZ63-4S/9311 and Y58S/9311. Test of BPH response showed that lines with three gene combined were more resistant than lines with two genes or lines with single genes.
The three-line hybrid breeding system is an important method for producing commercial rice hybrids in China and elsewhere. The system involves a cytoplasmic male sterile (CMS) line (A), a fertile maintainer line (B), and a restorer line. The A line is used as a female parent and the restorer line is used as male parent for hybrid seed production. The B line is near-isogenic to the A-line except for its pollen fertility. The B line has viable pollen grains and normal seed setting, so it can pollinate the A line, while the F 1 plants from this pollination are male sterile, again. In this way, the male sterility of the A line is maintained, and the A line is reproduced for further use or commercial use in a large scale. Jin 23B is a maintainer line of male-sterile Jin 23A, whereas Jin 23A is an elite indica cytoplasmic male sterile (CMS) line widely used to produce three-line hybrids over the last 20 years in China. More than 100 different hybrids are derived from Jin 23A (http://www.ricedata.cn/), and in 2008 and 2009, Jin 23A-derived hybrid varieties are grown on about 1.7 million ha of rice in China with the highest ranking (Shi et al. 2012). However, these hybrids and their parents have become increasingly susceptible to blast and BPH due to absence of resistance genes and narrow genetic diversity. Blast resistance was recently improved by the introduction of genes Pi1, Pi2 and D12, into Jin 23B by marker assisted backcross breeding (Jiang et al. 2012). To similarly improve BPH resistance, we introduced BPH resistance genes Bph3, Bph14, Bph15, Bph18, Bph20 and Bph21 into Jin 23B with comprehensive evaluation. These lines will contribute to ensuring the security of rice production in China.
The BPH used for infestation were collected from rice fields at Wuhan, in 2012, and were maintained in a cage by continuous rearing on the susceptible variety TN1, at the glasshouse facility at Huazhong Agricultural University. Before use in experiments, one-to five-instar BPH nymphs were prepared and maintained on TN1.
The whole-genome single nucleotide polymorphism (SNP) array RICE6K was used for genetic background profiling of the introgression lines (Yu et al. 2014). Genomic DNA from each line was extracted from 10 dry seeds. DNA amplification, fragmentation, chip hybridization, single base extension, staining and scanning were conducted by the Life Science and Technology Center, China National Seed Group Co., LTD (Wuhan, China).

Evaluation of BPH response in seedlings and tillering stage of plants
An Experiment was conducted by using a modified bulk method under sophisticated greenhouse condition followed Pathak et al. (1969). PTB33 and TN1 were included as controls in all experiments. Lines were sown in plastic trays (60 cm length × 40 cm width × 10 cm height) with 12 plants per row, 11 rows per tray. When seedlings at 10 days post sowing (three-leaf stage) were infested with second and third instar nymphs (twelve nymphs per seedling). Each experiment was conducted in a randomized complete block design with six replications. Plant responses were recorded seven days after infestation (DAI) when all seedlings of TN1 were almost dead. BPH resistance was evaluated with the six-scale standard of scoring system, described by Huang et al. (2001): 0 = no damage; 1 = very slight damage; 3 = first and second leaves partially yellowing; 5 = pronounced yellowing and stunting; 7 = mostly wilting, the plant still alive; 9 = the plant completely wilted or died, in which higher scores indicate greater susceptibility to BPH.
The recurrent parent, backcross lines, resistant check and susceptible check (TN1) seeds were sown in the field at Wuhan for plant BPH response assessments. Fifteen days later two plants of each line and checks were transplanted into a plastic pot (16.5 cm of bottom diameter × 22.5 cm of top diameter × 19 cm of height) containing pulverized soil in ten replications. All the plants were grown in a cage at the glasshouse facility. At the tillering stage (30 days post transplanting) the plants were infested with second and third instar nymph at a density of 300 nymphs per plant. When the susceptible check was almost dead (approximately 17 days after infestation) all plants were assessed and scored as resistant (R), moderately resistant (MR), moderately susceptible (MS) and highly susceptible (S) (Suh et al. 2011). Scoring occurred a second time about 25 days post infestation.

Quantification of honeydew excretion from adult BPH females
Honeydew collection was accomplished under aseptic condition under greenhouse followed by Pathak et al. (1982) with minor modifications. Seven improved rice families with BPH resistance gene Bph3, Bph14, Bph15, Bph18, Bph20, Bph21 and Bph14/Bph15 and the control Jin 23B and PTB33 were planted in plastic pots. 50 days after transplanting, five-instar BPH female nymph (previously starved for 2 h) were inoculated with a suitable parafilm sheet (3.5 × 4.5 cm) (Incubation rate = two females per sheet) and subjected to the improved lines with suitable check. After 24 h of feeding, insects were removed from the sheet, and honeydew in each sheet were weighted using a 0.1 mg sensitivity balance. An experiment was repeated over four replications.

BPH survival on rice plants
BPH survival measurements were made on the backcross lines and suitable checks following the method of Sebastian et al. (1996). Individual seedlings were grown on a rice liquid culture medium in 500 ml volumetric flasks. Ten insects with second-instar nymph were subsequently put into each volumetric flask, and survivability of BPHs in each flask were recorded in every 2 days of interval for 10 days. The experiment was carried out in greenhouse, and the temperature was maintained at 25-28°C. BPH survival percentage was calculated as follows. BPH survival percentage = Number of surviving nymph divided by the total number of released nymph. This measurement was repeated for six times.

Pollen fertility of improved lines
Fertility assessments of the backcross lines were made using six to ten selected lines for each genotype. Crosses were made to the male sterile Jin 23A, and the F 1 plants were used as female parents in crosses with the corresponding improved lines. Six spikelets per line of each F 1 and BC 1 F 1 individual were sampled for assessment of pollen fertility using 1% I2-KI staining (Virmani et al. 1997). The fertility rate of each line was scored as completely sterile (0%), partially sterile (1-30%), partially fertile (> 30-60%) and fertile (> 60%) (Govinda Raj and Virmani 1988).

Evaluation of agronomic traits in the field
Agronomic assessments of the seven backcross lines were carried under natural field conditions, seven homozygous rice families and the control (Jin 23B) were planted in randomized complete block design in Wuhan, China in the autumn of 2013. A trial was performed with six replications, each plot consisted of three rows and each row consisted of 12 plants at a planting density of 17 × 27 cm (17 cm plant to plant and 27 cm row to row distance). Seven plants from central row of each plot were subjected to agronomic trait measurements, including days to heading (DTH) and plant height (PH). After the maturity of rice, the seven plants from central row were harvested individually for panicle number (PN), number of grains per panicle (NGP), spikelet fertility (SF), weight of 1000-grains (GW) and yield per plant (YD) measurement. Filed management followed normal agricultural practices with the nitrogen, phosphorus, potassium fertilizer of 180, 65, and 65 kg/hm 2 respectively.

Data analysis
The available RM marker series were searched in the rice genomic database (www.gramene.org). InDel markers were designed based on the references genome of Nipponbare and 9311 reference sequences. Date analysis was performed using one-way ANOVA and Tukey's tests in Microsoft Excel 2003 or SPSS 17.0.

BPH survival rates
To know the influence of antibiosis in the BPH resistance Bph3, Bph14, Bph15, Bph18, Bph20 and Bph21, the BPH survival rate was measured on the pyramided line and introgression lines in every two days interval for 12 days. As shown in Fig. 6, the average BPH survival rate on Jin 23B (14/15) was the lowest among all genotypes, which was 21.11% at 12 DAI. The average numbers of BPH surviving on Jin 23B (15), Jin 23B (14) and Jin 23B (21) decreased gradually and showed a significant difference in number compared with other genotypes at 8 DAI, and the average survival rates were 34.44%, 38.89% and 51.
Generally, these findings indicated that the BPHs probably had stronger effect on the antibiotic factors in the pyramiding line and introgression lines. And in the seven genotypes, the order of the gene effect on BPH survival at 8 DAI was: The maintainer ability of improved lines In order to evaluate the fertility of the introgression lines as maintainers we assessed the pollen viabilities of hybrids with Jin 23A by I2-KI staining. Of the nine improved lines with Bph3, one line showed complete fertility (81.7%). Among the eight Bph14 introgressed lines, two lines were completely fertile (94.0%) or partially fertile (53.0%). Among the eight Bph15 introgressed lines, two lines showed completely fertile (70.7%) or partially sterile (15.0%). Among nine Bph18 lines, two lines showed complete fertility (100% and 60.4%) and one line showed partially fertile (56.7%). Among ten Bph21 lines, one line showed partial sterility (16.7%) and one line showed complete fertility (100%). Among eight Bph14/ Bph15 lines, one line showed partially fertile (48.3%). For Bph20 introgressed lines, 6 lines were showed complete sterility (0%). The lines with fertile rate more than 0% in the F 1 population were excluded, and all the improved lines were showed complete sterility (0%) in the BC 1 F 1 population. Conclusion was that more than five lines of each genotype showed maintaining ability in the Jin 23B background.

Agronomic traits of improved lines of Jin 23B
We compared the agronomic traits of one improved line from each of the 7 backcross populations with Jin 23B at Wuhan under stress-free conditions in 2013 (Table 2). Improved lines of Jin 23B (14), Jin 23B (18) and Jin 23B (14/15) were significantly lower in PH; and Jin 23B (20) had a significant increase in PN (Table 2). In addition, two improved lines, Jin 23B (3) and Jin 23B (20) had significantly decrease in NGP and SF; One improved line, Jin 23B (3), had significantly increase in GW. Furthermore, the yields of improved lines were found similar to the control (Jin 23B) under natural field conditions (Table 2). Generally, these results showed that two improved lines Jin 23B (15) and Jin 23B (21) had similar agronomic trait relative to the control maintaining the elite agronomic traits.

Discussion
Mapping and cloning of functional genes in rice have facilitated MAS of genes for resistance to biotic stress in rice genetic improvement in recent years. Markerassisted selection combined with backcross breeding is a powerful method of coping with defects in otherwise very susceptible rice cultivars which had application in Photograph of resistance level to BPH of improved Jin 23B with resistance gene Bph3, Bph14, Bph15, Bph18, Bph20, Bph21 and Bph14/Bph15 at the seedling stage (the 7 days after infested with BPH). PTB33 (carrying Bph3) is a resistance control and TN1 is a susceptible control bacterial blight, sheath blight, blast, stem borer, and gall midge resistance (Huang et al. 1997;Katiyar et al. 2001;Datta et al. 2002;Maruthasalam et al. 2007;Koide et al. 2010;Jiang et al. 2012). For BPH resistance, recent studies have focused on incorporating BPH resistance genes such as Bph14, Bph15, and Bph18 into the elite rice cultivars Minhui 63, 9311 and their hybrids (Hu et al. 2012(Hu et al. , 2013. In the present study, six dominant BPH-resistance genes, Bph3, Bph14, Bph15, Bph18, Bph20 and Bph21 were introgressed or pyramided into an important hybrid parent Jin 23B using MAS-based backcrossing, and improved lines containing single or two BPHresistance genes were obtained. BPH bioassays showed that each of the six BPH resistance genes conferred different levels of resistance at both seedling and tillering stage of plants. The line with the combined Bph14 and Bph15 genes was more resistant than lines with the respective single genes and was the most resistant line overall. And the resistance level of single gene introgression line with Bph15 was significant higher than that with Bph3 or Bph18 at seedling stage. Pyramiding line or introgression lines carrying gene Bph14/Bph15, Bph15 or Bph14 had a moderately resistant level at 25 DAI at tillering stage. And single introgression lines carrying Bph20, Bph21, Bph3 or Bph18 had a moderately susceptible level at 25 DAI at tillering stage. In honeydew weight and BPH survival rate, the pyramiding line with two genes had a significant difference from single introgression lines. And the single introgression lines had different among themselves (Figs. 3b,6). The resistance levels of donor parent PTB33 at both seedling and tillering stages were significantly higher and the honeydew weight of PTB33 was significantly lower than Jin 23B (3), indicating that PTB33 may have more than one BPH resistance gene. In fact, Lakshminarayana and Khush (1976) reported that PTB33 carried two BPH resistance genes, namely bph2 and Bph3. From the resistance level of Jin 23B (3), we hypothesis that Jin 23B (3) did not get the bph2 gene from PTB33.
Gene effect of the BPH resistance conferred by Bph3, Bph14, Bph15, Bph18, Bph20, and Bph21 was almost the same in honeydew weight, BPH survival rate, and the tests in seedlings stage. The pyramiding line carrying Bph14 and Bph15 has higher resistance level than the lines carrying single resistance gene. The results indicated an additive effect of pyramiding genes and showed that pyramiding two resistance genes have larger effects to BPH under Jin 23B background. In addition, single gene introgression line carrying Bph15 had a largest effect than other single gene introgression lines. And single gene introgression line carrying Bph18 had a lowest effect than other single gene introgression lines.
With the development of high-throughput and resolution genotyping platform in whole genome scale, whole genome SNP array-based background selection has become a feasible strategy with higher precision and efficiency for crop genetic improvement. The breeding chip RICE6K was employed in background examination of the BPH resistance improved lines. After three times backcross, all the improved lines had the segment of the target gene from donors. The genetic background recovery rate of Jin 23B (18), Jin 23B (21) and Jin 23B (14/15) were less than 90%, and the other segments from the donor parents may have had resistance QTL with positive effects on BPH resistance. Further backcrosses with Jin 23B would eliminate those segments. In addition,  (20) and Jin 23B (21) did not get the chromosome 6 resistance QTL from the IR71033-121-15 donor. Eight BPH-resistance genes (Bph1, Bph2, Bph7, Bph9, Bph10, Bph18, Bph21, Bph26) that are clustered on the long arm of chromosome 12, and they may be allelic. Among them Bph26, Bph9 and Bph18 have been cloned. Zhao et al. (2016) reported that Bph9 shares 91.32% and 96.05% nucleotide sequence identity with Bph26 and Bph18, respectively, and these allelotypes confer varying levels of resistance to different biotypes of BPH. The coding region of BPH1/9 shows a high level of diversity in rice germplasm (Zhao et al. 2016). In our study the allele Bph18 and Bph21 were introduced into Jin 23B and the resistance level conferred by Bph21 was higher than Bph18 at seedling stage.
Most of all, to pursue a durable and broad-spectrum resistance to BPH we should not only make use of gene pyramiding, but also exploit genetic diversity as an ecological approach to pest control, which can be highly effective over a large area and contribute to the sustainability of crop production (Zhu et al. 2000). It is more important that diversity represents insurance of buffering from failure of a single widely used gene. Because different BPH resistance genes are identified from different indica and wild rice accessions. Using of improved lines containing one or both BPH resistance genes planted in the different or same fields should restrict rapid increases in insect population and therefore reduce crop damage. Fortunately, Bph14, Bph15 and Bph18 have been pyramided into several rice hybrids in different genetic backgrounds, indicating that planting resistant pyramided hybrids in conjunction with conventional susceptible hybrids could effectively decreases the BPH population with the consequence of less serious BPH outbreaks, reduced costs of labor and pesticides, and increased rice production (Hu et al. 2011). The improved lines generated in this study can be used to develop multiple gene pyramided lines in the same genetic background for BPH control in hybrids.  Jin 23B (Bph14/15) 62.7 ± 0.9 88.6 ± 5.3* 9.6 ± 3.2 98.7 ± 20.8 81.1 ± 5.6 24.2 ± 1.7 22.7 ± 8.0 Significantly different from the performance of the improved lines with Jin 23B at * P < 0.05 and ** P < 0.01 Data from filed trails demonstrated that most of the required traits were well recovered in the improved lines ( Table 2). The Bph3 and Bph20 introgression lines had a significantly lower NGP and SF relative to the control, but showed significantly increase in GW or PN. The yields of improved lines were similar to the control. Two improved lines Jin 23B (15) and Jin 23B (21) showed no difference in agronomic traits compared to Jin 23B. Jin 23A is an elite indica CMS line widely used in hybrid rice production in China. Through hybrid with Jin 23A, these BPH resistance gene can be transferred into Jin 23A, which can be used for development of BPH resistance hybrids in future. In addition, I2-KI stain of the F 1 and BC 1 F 1 hybrids show that the improved Jin 23B lines have good maintainer ability. These results show that these BPH resistance genes in Jin 23B have considerable commercial value for use in sustainable rice production.
From a global viewpoint, the increasingly severe occurrence of insects and diseases and indiscriminate pesticide is a challenge for sustainable rice production. A combination of approaches based on recent advances in genomic research has been formulated to address these challenges, with the long-term goal to develop rice cultivars referred to as Green Super Rice (Zhang 2007). Green Super Rice should possess resistance to multiple insect pests and diseases. In our previous work we have got the improved line Jin 23B (Pi1/Pi2/D12) containing three blast resistance gene (Jiang et al. 2012). This line can be hybridized with the present BPH resistance lines. The objective of the new improved lines is to address the problems of both blast and BPH.

Conclusions
Overall, six dominant BPH-resistance genes Bph3, Bph14, Bph15, Bph18, Bph20 and Bph21 have an effect on BPH growth and development and antibiotic factors, resulting both seedling and tillering resistance. Bioassay test showed that the order of the gene effect being Bph14/Bph15 > Bph15 ≥ Bph14 ≥ Bph20 ≥ Bph21 ≥ Bph3 > Bph18 > none at seedling stage of rice. Furthermore, field trial data demonstrated that yields of the improved lines were similar to the Jin 23B control under field conditions. These improved versions can be used in hybrid production in China.