Introduction

Rice black-streaked dwarf disease (RBSDD) is a plant disease caused by rice black-streaked dwarf virus (RBSDV), which is transmitted by small brown planthopper (SBPH, Laodelphax striatellus Fallén) in a persistent propagative manner (Van et al. 2000). Rice black-streaked dwarf virus is a double-stranded RNA (dsRNA) virus belonging to the second group of the genus Fijivirus, and the disease it causes is easily transmitted and can lead to epidemics (Lee et al. 2005). In addition to rice, it can also infect wheat, barley, corn, and other gramineous crops, and causes serious damage to crop production (Bai et al. 2002). Typical symptoms of plants infected with RBSDV include stunted growth, dark green leaves, white waxy or black-streaked enations on the abaxial surface of leaves, and poor heads or no heads at all. In rice production, spraying chemical pesticides is still the main method of controlling SBPH and RBSDD, but this has proven ineffective with recent increases in SBPH migration and outbreak frequency. In addition, pesticides are expensive and harmful to the environment, and can also enable SBPH to readily develop resistance (Liu et al. 2020). Therefore, exploring new rice genes from diverse sources and developing new resistant cultivars are considered to be the most economical and effective strategies for SBPH and RBSDD management (Wang et al. 2014; Vicheka et al. 2017).

The Zhejiang Province of China suffered an outbreak of RBSDD in the 1960s, but since then, RBSDD had entered a period of dormancy, therefore failing to arouse the attention of breeders and farmers. Thus, screening of germplasms conferring RBSDD resistance has begun only recently, few varieties and QTL with high RBSDD resistance level have been identified to date, and the molecular mechanism underlying RBSDD resistance remains poorly understood. Using a set of chromosome segment substitution lines (CSSLs) by backcross between Zhenshan 97 and Minghui 63, a major QTL, qRBSDV-6MH, was fine-mapped at a physical distance of 627.6 kb on chromosome 6 (Li et al. 2013). In our previous study, three QTL for RBSDD resistance located on chromosomes 6, 9, and 11, were repeatedly detected by using an F2:3 population derived from a cross between 9194 and Suyunuo (Sun et al. 2017). Xu et al. (2018) identified five QTL for RBSDD resistance, which were distributed on chromosomes 3, 6, 7, 9, and 11, through nature infection in the field of RBSDD resistance of an F2:3 population derived from WR24 and Suyunuo. In addition to bi-parental QTL analysis, genome-wide association study (GWAS) has become a popular strategy for mining QTL in plants. Xiao et al. (2019) identified a significant RBSDD-associated locus, named as qRBSDVD6, by GWAS analysis, and the QTL was confirmed by a recombinant inbred line (RIL) population derived from W5686 and Suyunuo. In another study conducted by Feng et al. (2019), GWAS on RBSDD resistance was conducted in the field using the rice diversity panel 1 (RDP1) cultivars, two QTL, qRBSDV-4.2 and qRBSDV-6.3, were detected in two environment conditions, furthermore, introgression of qRBSDV-6.3 into two susceptible rice cultivars by marker-assisted selection (MAS) demonstrated the effectiveness of qRBSDV-6.3 in enhancing RBSDV resistance. Wang et al. (2022) identified a highly resistant variety W44 through extensive evaluation of a diverse international rice panel, using genome-wide association study, a significant associated locus with largest phenotypic effect, namely, qRBSDV6-1, was finely mapped to a 0.8–1.2 Mb region on the short arm of chromosome 6, and the candidate gene of qRBSDV6-1, aspartic proteinase 47 (OsAP47), was identified and functionally confirmed as the first cloned gene for its resistance to RBSDD in rice, recently. Liu et al. (2021) identified a significant RBSDD-associated locus, named as qRBSDV-6.1, using a different population by multi-locus GWAS, haplotype analyses of the candidate genes LOC_Os06g03150 suggested that the gene is mainly associated with the differentiation of resistance within the xian subgroup, meaning this gene is very important for underlying RBSDD resistance to xian accessions. In addition, the interaction between SBPH, rice plants, and RBSDV is complex and diverse, so there is an urgent need to map more QTL in order to breed resistant cultivars to reduce the damage caused by RBSDD (Zhou et al. 2018).

In order to identify reliable and highly resistant germplasms and QTL for RBSDD resistance, 248 germplasm resources collected from different countries and regions were identified by field identification in our study. Among them, 19 rice varieties (including H185) were highly resistant to RBSDD. An F2:3 population of H185/WYJ3 was then constructed to detect QTL with RBSDD resistance, and two QTL, qRBSDD2 and qRBSDD7, were repeatedly detected in artificial inoculation conditions, accounting for 57.8% of the total phenotypic variation. Furthermore, introgression of qRBSDD2 and qRBSDD7 into a susceptible rice variety by MAS demonstrated that those two QTL are reliable and can act additively. The results of this study will provide useful genetic resources and molecular markers for breeding new cultivars conferring resistance to RBSDD.

Materials and methods

Plant materials

A total of 248 rice varieties collected from different rice-growing regions were selected to evaluate RBSDD resistance by natural infection in the field in 2015 and were provided by the Institute of Germplasm Resources and Biotechnology, Jiangsu Academy of Agricultural Sciences, China. An F2 population containing 188 individuals was constructed by crossing H185 (indica) and WYJ3 (japonica), and then, each F2 plant was self-crossed to generate the corresponding F2:3 line. 188 F2:3 individuals and the two parental varieties were used for field and artificial identification in 2016 and 2017, respectively. Among them, 133 and 83 individuals with good repeatability were selected to detect QTL related to RBSDD. The indica cultivar 9194 with high resistance to RBSDD was used as a resistant control, while the japonica rice cultivar Suyunuo (SYN) was used as a susceptible control, as described in Sun et al. (2017).

Evaluation of RBSDD resistance

In 2015–2016, experimental fields in Kaifeng City in Henan Province were chosen for field natural evaluation, due to the prevalence of RBSDD in this region. Field and artificial inoculation identification of RBSDD were carried out as described in our previous study (Sun et al. 2017). Briefly, about 15 days before harvesting winter wheat, 60 seeds per cultivar or F2:3 individual were sown in a rice seedling bed encircled by wheat plants infected with RBSDV. At 4 weeks after sowing seeds, 40 seedlings per cultivar or F2:3 individual were transplanted into an experimental area at Kaifeng, and the plants were cultured with normal field management, except that pesticides were not applied during the rice-growing period. The plants with the typical symptoms of RBSDD were considered to be susceptible plants. The incidence of RBSDD was recorded at the peak tillering stage as the infected plants showed obvious and typical disease symptoms at this stage. Each cultivar or F2:3 individual was replicated three times, and the average disease incidence (DI, the number of RBSDV-infected plants/the total number of plants counted × 100%) was used for QTL analysis.

Linkage mapping and QTL analysis

The polymorphisms of H185 and WYJ3 were detected by 412 pairs of simple sequence repeats (SSR) and 124 pairs of insertion/deletion (InDel) markers evenly distributed across 12 rice chromosomes. Then, 133 F2:3 individuals of H185/WYJ3 were analyzed by 142 pairs of markers with polymorphic differences among the parent plants. The precise names of SSR marker and sequences of InDel marker are shown in the Supplementary materials. Linkage groups were assembled using Mapmaker/EXP 3.0 (MapMaker3.0b for Windows http://www.softpedia.com/get/Science-CAD/MapMaker.shtml) (Lander et al. 1987). The total length of the map was 1426.3 cM, and the average distance between markers was 10.04 cM, which was suitable for QTL analysis. Following this, QTL analyses of F2:3 population containing 133 or 83 plants was carried out by composite interval mapping using QTL Cartographer, version 2.5 (2010 N.C. State University, Bioinformatics Research Center, Raleigh, NC, USA). The percentage of phenotypic variation explained (PVE) for each QTL and the additive effects were also obtained using the software.

Introgressing and pyramiding qRBSDD2, qRBSDD7 through marker-assisted selection (MAS)

To introgress and pyramid RBSDD resistance QTL to WYJ3, we firstly crossed H185 with WYJ3 and then resultant hybrid F1s were continuously backcrossed with WYJ3 five times. In each generation, two flanked markers of each QTL were deployed to select target recombinants. Finally, eight none, single or two QTL introgressed homozygous plants selected from the BC5F2 population were identified by artificial inoculation, each line had five replicates, and each replicate had 40 plants. SNP genotyping was performed by Huazhi Biotechnology Company using multiplex PCR with next-generation sequencing on the high-throughput genotyping platform, lllumina X-10.

Results

Screening of RBSDD-resistant germplasm resources

In this study, 248 germplasm resources from different countries and regions were screened for resistance to RBSDD (Fig. 1A). The results showed that the disease incidence (DI) of the 9194 (resistant control) was 9.2%, indicating high resistance to RBSDD, and the DI of the Suyunuo (SYN) (susceptible control) was 82.6%. The overall phenotype of all the tested accessions showed a left bias normal distribution, 19 varieties were highly resistant to RBSDD (DI < 10%), and 45 varieties were susceptible to RBSDD with DI of more than 50% (Fig. 1B). Among the 19 varieties with high resistance to RBSDD, 42.1% were japonica accessions and 57.9% were indica accessions. Among the 45 highly susceptible varieties, japonica and indica rice cultivars accounted for 64.4% and 35.6%, respectively. Among these, the indica cultivar H185 was highly resistant, while the japonica cultivar WYJ3 was highly susceptible (Fig. 1B). According to the analysis of the indica and japonica subsets, the DI of indica varieties was significantly lower than those of japonica varieties (Fig. 1C). Based on the comparison of the DI of different germplasm resources in different countries, the average DI of indica varieties was lower than that of japonica varieties in all countries except Thailand (Fig. S1A), and the same result was found across the different provinces of China (Fig. S1B).

Fig. 1
figure 1

Geographical distribution and evaluation for rice black-streaked dwarf disease (RBSDD) resistance of all the tested rice accessions. A Geographical distribution of 248 rice species in various countries and provinces of China; B frequency distribution of 248 germplasm resources for resistance to RBSDD; C distribution of RBSDD incidence rate of germplasm resources with indica and japonica subgroups

Full size image

Resistance of H185 to RBSDD

In order to detect QTL conferring resistance to RBSDD, H185 and WYJ3 plants were identified by field identification and artificial inoculation in 2016 and 2017, respectively. Using field identification, the average DI values of H185 and WYJ3 were 8.9% and 58.3%, respectively (Fig. 2A, B, D). Using artificial inoculation, the average DI values of H185 and WYJ3 were 12.4% and 75.1%, respectively, (Fig. 2C, D). There was a significant difference between H185 and WYJ3, meaning that these two cultivars were ideal for RBSDD research.

Fig. 2
figure 2

Phenotypes of H185 and Wuyujing 3 (WYJ3) for rice black-streaked dwarf disease (RBSDD) resistance under natural and artificial inoculation conditions. A and B Representative image of H185 and WYJ3 inoculated with RBSDV in the natural field condition in 2016. C Representative image of H185 and WYJ3 artificially inoculated with RBSDV in 2017. D The disease incidence of H185 and WYJ3 by the field identification and artificial inoculation. n = 3, **p < 0.01 by Student’s t-test

Full size image

Genetic analysis and identification of QTL for resistance to RBSDD

To detect QTL for RBSDD resistance, an F2 population including 188 F2:3 individuals derived from H185 and WYJ3 was used in natural infection experiments in 2016. Of the 188 individuals, 133 with good repeatability (less than 5% variation) were selected to further detect QTL for RBSDD resistance. The DI frequency distribution of the 133 F2:3 individuals varied continuously, ranging from 0 to 61.9%, and there was a peak at 20%, which exhibited characteristics of quantitative traits (Fig. 3A). Those results suggested that resistance to RBSDD was controlled by multiple QTL in this population.

Fig. 3
figure 3

Distribution of RBSDD resistance scales of the F2:3 families under natural (A) and artificial (B) inoculation conditions

Full size image

Based on the field phenotypic data of the 133 F2:3 individuals, the RBSDD resistance QTL in H185 were detected by complex interval mapping using the Windows QTL Cartographer V2.5 software. As shown in Table 1 and Fig. 4, three QTL, namely qRBSDD2, qRBSDD7, and qRBSDD11, were detected on chromosomes 2, 7, and 11, respectively. These QTL were located between the markers RM425-RM208, RM3484-RM6776, and RM26105-Indel11-3, respectively. They explained 20.8%, 21.4%, and 11.4% of the phenotypic variance, respectively, with LOD scores of 5.35, 4.81, and 2.65, respectively (Table 1). Jointly, they explained 53.6% of the total phenotypic variance in this population. The additive effect analysis showed that the above-mentioned three QTL were all inherited from the resistant cultivar H185.

Table 1 Quantitative trait loci (QTL) for rice black-streaked dwarf disease (RBSDD) resistance detected in the H185/Wuyujing 3 (WYJ3) F2:3 population
Full size table
Fig. 4
figure 4

Chromosome location of QTL for RBSDD resistance in H185/WYJ3 F2:3 population. Marker names are given on the right of the chromosome, and map distances (cM) on the left

Full size image

Verification of QTL resistant to RBSDD

In order to verify the three QTL resistant to RBSDD in the present study, the two parental cultivars and 83 F2:3 individuals were identified by artificial inoculation in 2017 (Fig. 3B). Using a QTL mapping strategy, two QTL, qRBSDD2 and qRBSDD7, could be re-detected and stably expressed, the LOD scores were both greater than 4, and the phenotypic variance were more than 25%, which is thought to indicate major QTL (Table 1 and Fig. 4). Taken together, these data further confirmed that these QTL were reliable.

Pyramiding effect of QTL on RBSDD resistance

The pyramiding effect of two stable QTL, qRBSDD2 and qRBSDD7, was analyzed by comparing the mean DI of different lines from a BC5F2 population by the backcross between H185 and WYJ3, with WYJ3 as the recurrent parent. The genotype of eight BC5F2 individuals was assessed using a PCR-based genotyping protocol, indicating the proportion of WYJ3 genotype was more than 95% (95.2–97.7%), and none of the BC5F2 lines carried the resistant allele of qRBSDD11 detected in our study. As shown in Fig. 5, two lines carrying none of the QTL showed a susceptible phenotype with an average DI of 67.5% and 63.8%, respectively, while those introgressed lines possessing any single QTL of qRBSDD2 or qRBSDD7 exhibited moderate resistance, with DI ranging from 36.2% to 43.7%. Remarkably, two QTL pyramided lines showed higher resistance for RBSDD than other lines, with an average DI of 21.4% and 22.9%, respectively (Fig. 5). Taken together, these data further confirmed that these two QTL are reliable and can act additively.

Fig. 5
figure 5

Pyramiding effect of two stable QTL associated with resistance to rice black-streaked dwarf disease (RBSDD). The qRBSDD2, qRBSDD7, or two loci pyramided lines, selected from a BC5F2 population by the backcross between donor parent H185 and recurrent parent WYJ3, were identified by artificial inoculation, n = 5. Bars denoted by different letters are significantly different at p < 0.05

Full size image

Discussion

With global warming and changes to the cultivation system, the outbreak of RBSDD has posed a serious threat to rice production in Asia regions, especially in eastern China. There have been many reports on the symptoms of the disease, but due to complex epidemiological factors, cumbersome field experiments, the need for large-scale investigations, and a longer research cycle, the systematic and in-depth study of RBSDD has become extremely difficult (Li et al. 2013). Therefore, little research has been conducted on the inheritance of or breeding strategies that could be used to enhance RBSDD resistance, and no rice variety resistant to RBSDD has been widely used in rice production. It is worth noting that the average DI of japonica varieties in the Philippines and Indonesia was greater than 50% (Fig. S1A). In China, the resistance of indica varieties in Hubei Province was very weak, with an average DI of more than 60% (Fig. S1B). These countries or regions seriously lack RBSDD-resistant germplasm resources, and the consequences may be very serious once disease outbreaks occur. Therefore, it is urgent to accelerate the screening of rice germplasm resources in order to map major QTL to breed new cultivars with high resistance to RBSDD.

Because resistance mechanisms against insect-borne viruses are very complex, phenotypic identification of viral diseases has always been an issue (Mandadi and Scholthof 2013; Souza and Carvalho 2019). Rice black-streaked dwarf disease is transmitted by SBPH in a persistent manner, but RBSDV cannot be transmitted to the next generation via eggs, which makes artificial inoculation and identification of RBSDD extremely difficult (Xiao et al. 2019). In addition, field natural identification is convenient and easy to implement. Large-scale identification of RBSDD relies mainly on field natural identification in severely diseased areas. However, owing to the uneven distribution of vector insects, the accuracy of field natural identification needs to be further improved (Zheng et al. 2012). In order to improve the reliability of field natural evaluation, the following measures were taken in our study: (A) Selecting regions with serious disease outbreaks as identification nurseries on the premise that the DI of susceptible varieties was more than 50% in the past year, and eventually, experimental fields in Kaifeng City in Henan Province were chosen for field natural evaluations due to the prevalence of RBSDD in this region. (B) Increasing the SBPH population: Experimental fields were surrounded by wheat fields, after wheat harvest, a large population of SBPH moved to the experimental field to ensure the quantity of vector insects. (C) Improved uniformity of insect exposure: After the wheat harvest, the SBPH were expelled several times a day in the seedling field. (D) Evaluation of RBSDD resistance by artificial inoculation: RBSDV-carrying SBPH individuals were used to inoculate the F2:3 population, which ensured the uniformity and adequacy of inoculation intensity. The above measures ensured the reliability of the phenotypic evaluation.

The present study found that the resistant parents, such as Minghui63, Tetep, WR24, and, W5686, were all indica varieties in the previous QTL mapping of RBSDD resistance (Li et al. 2013; Zhou et al. 2015; Xu et al. 2018; Xiao et al. 2019). Similarly, H185, the resistant parent in our study, was also an indica variety, which indicated that indica rice varieties were more resistant to RBSDD than japonica rice in the evolutionary process. This hypothesis was further confirmed by comparing the DI of indica and japonica subsets in different countries and regions.

Previous studies have shown that resistance in the F2 population obtained by crossing resistant and susceptible germplasms shows the characteristics of a normal distribution (Zhou et al. 2015; Zhang et al. 2016; Xu et al. 2018), meaning that resistance of RBSDD is controlled by multiple QTL. These genetic characteristics have been confirmed in maize rough dwarf disease (MRDD), which was also caused by RBSDV (Luan et al. 2012; Shi et al. 2012; Tao et al. 2013). In this study, three RBSDD-resistant QTL were also detected in the F2:3 population of H185/WYJ3, which supports the results of previous studies. These results provide useful information for enhancing the breeding of new varieties resistant to RBSDD, and lay a foundation for exploring the molecular mechanism behind RBSDD resistance.

Among the three QTL detected in our study, qRBSDD2 and qRBSDD7 explained 20.8% and 21.4% of the phenotypic variance, respectively, with LOD scores of 5.35 and 4.81, respectively, suggesting that they are major QTL involved in resistance to RBSDD. In order to further confirm the existence of these two QTL, 83 F2:3 individuals of H185/WYJ3 were re-identified by artificial inoculation, the presence of these two QTL was re-detected at the same location, which indicated that qRBSDD2 and qRBSDD7 were stably expressed major QTL for RBSDD resistance. Marker-assisted pyramiding of QTL is an effective way to improve quantitative traits. In this study, we determined the pyramiding effect of these two QTL in BC5F2 population derived from the backcross between donor parent H185 and recurrent parent WYJ3, and the pyramiding lines of the two QTL exhibited higher resistance for RBSDD than lines with no or a single QTL. Therefore, qRBSDD2 and qRBSDD7 have great potential for enhancing the breeding of new rice varieties resistant to RBSDD. Two tightly linked markers with the two QTL (SR2-4 and SR7-3, Supplementary materials) were designed and can be used in breeding cultivars with RBSDD resistance by marker-assisted selection (MAS).

At present, approximately 20 QTL conferring resistance to RBSDD have been detected on different rice chromosomes. In previous studies, Zheng et al. (2012) detected a RBSDD-resistant QTL named qRBSDV2 on the short arm of chromosome 2, while the qRBSDD2 detected in the present study is located on the long arm of chromosome 2. Xu et al. (2018) detected a RBSDD-resistant QTL named qRBSDV7WR24 on chromosome 7, while according to the physical locations in the rice genomes, qRBSDV7WR24 and qRBSDD7 detected in this study are quite far apart, hence, qRBSDD2 and qRBSDD7 may be novel QTL involved in RBSDD resistance. Generally, chromosome 11 is considered to be a "hotspot" of rice resistance genes to many diseases, such as rice blast, bacterial blight, and stripe blight (Chauhan et al. 2002; Sallaud et al. 2003; Xiang et al. 2006; Zhang et al. 2011). In our previous study, a QTL conferring resistance to RBSDD, named as qRBSDV11, was detected on chromosome 11 using the F2:3 segregation population of 9194/Suyunuo, which explained 31.1% of the phenotypic variance with a LOD score of 7.16 (Sun et al. 2017). The chromosomal location of this QTL overlaps with that of the qRBSDD11 identified in this study, and these two QTL are likely controlled by the same gene or allele. The results indicate that qRBSDD11 is stable in different backgrounds and is valuable for MAS in breeding cultivars resistant to RBSDD.