Resistance architecture and breeding implications of blast and bacterial blight resistance genes in core WDR parents Authors: Zhening Miao, Jijin Chen, Deyan Kong, Jiahong Chen, Ziming Li, Xinqiao Yu, Hui Xia, Lijun Luo, Yi Liu Published: 2026 ABSTRACT BackgroundWater-saving and drought-resistance rice (WDR) is crucial for sustainable cultivation under water scarcity, yet its productivity is increasingly compromised by rice blast and bacterial blight (BB). A systematic assessment of resistance gene deployment in elite WDR parents is urgently needed to guide resistance breeding.ResultsWe dissected the resistance architecture of 47 core WDR parental lines by genotyping 11 blast and 4 BB resistance genes. For blast, Ptr was nearly fixed (97.9%), whereas loci previously reported to confer broad-spectrum blast resistance, such as Pigm (0%) and Pi9 (19.1%), were scarce. Blast resistance scores correlated negatively with gene number (ρ = -0.330, P = 0.023). For BB, resistance genes were only detected in 34.0% of lines, mainly restorers, with Xa23 (21.3%) and Xa7 (23.4%) being most frequent. The recessive gene xa5 was absent, and Xa21 (8.5%) was rare. BB resistance exhibited a strong negative correlation with gene count (ρ = -0.771, P < 0.001).ConclusionsResistance-associated loci were unevenly distributed in the WDR breeding panel, with apparent gaps in loci previously associated with durable or broad-spectrum resistance. These findings provide a preliminary basis for genotype-guided parental selection and marker-assisted pyramiding strategies in WDR breeding. Previous resistance-oriented improvement of elite restorers further suggests the potential value of integrating resistance-locus profiling with breeding design.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09285-0. Background Rice (Oryza sativa L.) sustains more than half of the world’s population. Climate change is increasingly destabilizing rice production by intensifying drought and flooding events, thereby weakening yield stability and increasing the likelihood of disease outbreaks [1]. Among biotic stresses, rice blast and BB are particularly destructive and can cause substantial yield losses during epidemics [2, 3]. As one of the most damaging fungal diseases of rice, blast—caused by Magnaporthe oryzae—remains challenging to control because rapid pathogen evolution often compromises the durability of single-gene resistance under field conditions. To date, 146 blast resistance (R) genes have been identified, among which 41 have been cloned and functionally validated [4, 5]. In breeding practice, broad-spectrum loci have been preferentially utilized, notably the Pi9/Pi2/Pigm cluster on chromosome 6 and allelic variants at the Pik locus on chromosome 11 [6, 7]. Beyond canonical NLR-mediated resistance, Ptr on chromosome 12 represents a non-NLR resistance gene [8], whereas Pi5 encodes a typical NBS–LRR protein; both have served as valuable donors for resistance improvement and gene pyramiding [9, 10]. As a widespread and destructive bacterial disease of rice, BB, caused by Xanthomonas oryzae pv. oryzae, threatens yield stability across diverse rice-growing ecosystems. More than 46 genes conferring resistance to BB have been reported, and 11 have been cloned to date [11]. Among these, two extensively deployed loci on chromosome 11, Xa21 and Xa23, provide broad-spectrum resistance. Notably, Xa23, derived from wild rice, is a single dominant gene that confers effective protection across developmental stages [12]. Xa7 is located on chromosome 6 and shows particularly strong resistance under high-temperature and high-humidity conditions [13]. The recessive gene xa5 on chromosome 5 is also effective against bacterial leaf streak [14]. Therefore, the genes selected in this study were not intended to cover all reported resistance loci, but were chosen to represent major loci with clear breeding relevance, available diagnostic markers, and practical value for resistance deployment and pyramiding in elite WDR parental lines. Collectively, the growing number of molecularly cloned resistance genes has made marker-assisted introgression and gene pyramiding increasingly feasible in practical breeding. However, durable field resistance cannot be inferred simply from the number of resistance genes deployed. Rapid shifts in pathogen populations can erode the effectiveness of single loci [15, 16], and genetic background effects or non-additive interactions may constrain the assembly and stable expression of optimal gene combinations [17, 18]. Therefore, the major bottleneck lies not in the availability of effective resistance loci, but in the rational deployment and targeted pyramiding of complementary loci in elite parental backgrounds to improve broad-spectrum and durable resistance. WDR combines the high-yield and high-quality attributes of irrigated rice with the water-saving and drought-tolerance advantages of upland rice [19]. When integrated with dry direct seeding and dry management, WDR cultivation can extend rice production to non-traditional areas, such as drylands and sloping fields, while improving resource-use efficiency and production cost-effectiveness. However, under increasing climate variability, stable resistance to blast and BB in WDR production cannot be assumed, particularly when core parental lines lack complementary resistance alleles [20]. Therefore, systematically characterizing the distribution and deployment of major resistance genes or target alleles in core WDR parents is essential for evidence-based parental selection and optimized pyramiding. This will help define priorities for parental resistance improvement and support the development of green, high-quality WDR varieties with improved disease resistance and yield stability across target production regions [21]. Methods Rice materials A total of 47 core WDR parental lines were used in this study, including 7 maintainer lines, 26 restorer lines, and 14 conventional inbred lines (Table S6). The maintainer lines were mainly derived from the Huhan1B background, represented by Huhan5B and Huhan7B. The conventional inbred lines were primarily developed from the Huhan15 background, including Huhan19, Huhan1509, Huhan1516, and Huhan1517. The restorer lines were mainly derived from the Hanhui3 background, including Hanhui9, Hanhui37, Hanhui49, Hanhui303, Hanhui8200, and Hanhui8209. These materials represent the major parental types and breeding backgrounds currently used in WDR improvement. Field evaluations were conducted across multiple site-years from 2024 to 2025 at the Zhuanghang Experimental Station in Shanghai, China, and the Lingshui Base in Hainan, China. At each site-year, the experiment was arranged in a randomized complete block design with three replications, and genotypes were randomly assigned within each block. For BB resistance markers, the monogenic lines IRBB5, IRBB7, IRBB21, and CBB23 were used as positive genotype references for xa5, Xa7, Xa21, and Xa23, respectively. For rice blast resistance markers, gene-specific donor lines or internal positive genotype references were included for each target locus: Pik, Hanhui49; Pib, Huhan7B; Ptr, Hanhui49; Pikh, Hanhui3015; Pikm, Huhan7B; Pita, Hanhui49; Pi9, Hanhui303; Pi5, Hanhui49; Pia, Hanhui3015; Pi2, Huhan1516; and Pigm, Gumei4. The blast resistance markers were functional or allele-specific markers located within target genes or closely associated with diagnostic variants. Marker calls were primarily determined according to the expected target amplicons and the corresponding positive genotype references. The internal WDR parental references were selected based on long-term marker-assisted selection and parental improvement practice. DNA isolation and DNA marker analysis Genomic DNA was extracted from 100 mg of young leaf tissue from each accession using a modified cetyltrimethylammonium bromide (CTAB) method [22]. DNA quality and concentration were assessed as described by Imam et al. (2014). Eleven tightly linked or functional markers (Table 1) were used to detect target marker alleles or haplotypes associated with 11 major rice blast resistance genes, namely Pik, Pib, Ptr, Pikh, Pikm, Pita, Pi9, Pi5, Pia, Pi2, and Pigm, and four BB resistance genes, namely xa5, Xa7, Xa21, and Xa23. All primers were synthesized by BioSune Biotechnology (Shanghai) Co., Ltd. Table 1Molecular markers used for detecting bacterial blight and blast resistance loci in 47 core WDR parental linesGeneChrMarkerPrimer sequence(5’-3’)Type of markerExpected size(bp)Reference Pib 2Pibdom-FGAACAATGCCCAAACTTGAGAFunctional marker365[23]Pibdom-RGGGTCCACATGTCAGTGAGC Pita 12YL155/YL87-FAGCAGGTTATAAGCTAGGCCFunctional marker1042[24]YL155/YL87-RCTACCAACAAGTTCATCAAA Pi2 6AP22 FGTGCATGAGTCCAGCTCAAALinked marker143[25]AP22 RGTGTACTCCCATGGCTGCTC Pi9 69-Pro-FTGATTATGTTTTTTATGTGGGGFunctional marker128[26]9-Pro-RATTAGTGAGATCCATTGTTCC Pigm 6Pigm-4-FATGCTCGATTCGTTACATTTFunctional marker82[27]Pigm-4-RCGTCCCACACTTTCTTTTT Pik 11Pik39512-FGCCACATCAATGGCTACAACGTTFunctional marker112[28]Pik39512-RCCAGAATTTACAGGCTCTGG Pikm 11Dkm1-FCTGGAGAGCTTCCGTGTCGACFunctional marker223[29]Dkm1-RTCTTCACGACGTCAATGGTGGC Pikh 11FMl43-FCCCAACATTGGTAGTAGTGCFunctional marker258[30]FMl43-RTCCTTCATACGCAACAATCT Pia 11PIA-FGCGACTGACACTTTCAATAGCFunctional marker189[31]PIA-RCGGTAGAGCAATTTAGAAGCAG Pi5 9PI5-ZW-FCCAAGTGCAACTAGAGGTATGGTFunctional marker1105[32]PI5-ZW-RGTGCATCATCTTCAGATATCAGG Ptr 12Z12-FTGCAGATTTGACTGCTCGGTFunctional marker200[8]Z12-RGGGATCTTCCTCGCCCAAA xa5 5xa5-FGCTCGCGTTCGTTCGTTAACFunctional marker295[33]xa5-RCCACAGATACCTTATCAAACAGCT Xa7 6Xa7-FAATATATAACCCCCCCCCCCCCAGFunctional marker482/342[33]Xa7-RTTAATTGCCACCGATGAGGTAATCC Xa23 11RM206-FCCCATGCGTTTAACTATTCTLinked marker120[34]RM206-RCGTTCCATCGATCCGTATGG Xa21 11pTA248-FAGACGCGGAAGGGTGGTTCCCGGALinked marker1000[35]pTA248-RAGACGCGGTGTAATCGAAAGATGAAA Table 1Molecular markers used for detecting bacterial blight and blast resistance loci in 47 core WDR parental linesGeneChrMarkerPrimer sequence(5’-3’)Type of markerExpected size(bp)Reference Pib 2Pibdom-FGAACAATGCCCAAACTTGAGAFunctional marker365[23]Pibdom-RGGGTCCACATGTCAGTGAGC Pita 12YL155/YL87-FAGCAGGTTATAAGCTAGGCCFunctional marker1042[24]YL155/YL87-RCTACCAACAAGTTCATCAAA Pi2 6AP22 FGTGCATGAGTCCAGCTCAAALinked marker143[25]AP22 RGTGTACTCCCATGGCTGCTC Pi9 69-Pro-FTGATTATGTTTTTTATGTGGGGFunctional marker128[26]9-Pro-RATTAGTGAGATCCATTGTTCC Pigm 6Pigm-4-FATGCTCGATTCGTTACATTTFunctional marker82[27]Pigm-4-RCGTCCCACACTTTCTTTTT Pik 11Pik39512-FGCCACATCAATGGCTACAACGTTFunctional marker112[28]Pik39512-RCCAGAATTTACAGGCTCTGG Pikm 11Dkm1-FCTGGAGAGCTTCCGTGTCGACFunctional marker223[29]Dkm1-RTCTTCACGACGTCAATGGTGGC Pikh 11FMl43-FCCCAACATTGGTAGTAGTGCFunctional marker258[30]FMl43-RTCCTTCATACGCAACAATCT Pia 11PIA-FGCGACTGACACTTTCAATAGCFunctional marker189[31]PIA-RCGGTAGAGCAATTTAGAAGCAG Pi5 9PI5-ZW-FCCAAGTGCAACTAGAGGTATGGTFunctional marker1105[32]PI5-ZW-RGTGCATCATCTTCAGATATCAGG Ptr 12Z12-FTGCAGATTTGACTGCTCGGTFunctional marker200[8]Z12-RGGGATCTTCCTCGCCCAAA xa5 5xa5-FGCTCGCGTTCGTTCGTTAACFunctional marker295[33]xa5-RCCACAGATACCTTATCAAACAGCT Xa7 6Xa7-FAATATATAACCCCCCCCCCCCCAGFunctional marker482/342[33]Xa7-RTTAATTGCCACCGATGAGGTAATCC Xa23 11RM206-FCCCATGCGTTTAACTATTCTLinked marker120[34]RM206-RCGTTCCATCGATCCGTATGG Xa21 11pTA248-FAGACGCGGAAGGGTGGTTCCCGGALinked marker1000[35]pTA248-RAGACGCGGTGTAATCGAAAGATGAAA Molecular markers used for detecting bacterial blight and blast resistance loci in 47 core WDR parental lines PCR amplification was performed in a 25 µL reaction volume containing 2.5 µL of 10× LA PCR Buffer (Mg²⁺ Plus), 4.0 µL of dNTP mixture (2.5 mmol L⁻¹ each), 1.0 µL of each primer (10 µmol L⁻¹), 50 ng of genomic DNA, 0.25 µL of TaKaRa LA Taq DNA polymerase (5 U µL⁻¹; Takara Bio Inc., Shanghai, China), and nuclease-free water to a final volume of 25 µL. The PCR program consisted of an initial denaturation at 94 °C for 5 min; 35 cycles of denaturation at 94 °C for 30 s, primer-specific annealing for 45 s, and extension at 72 °C for 2 min; followed by a final extension at 72 °C for 5 min. Each PCR-based marker assay was performed in three technical replicates using the same genomic DNA extract to confirm reproducibility. PCR products were separated on 2%–3% (w/v) agarose gels in 1× TAE buffer at 150 V for 90–120 min. DNA bands were visualized by ethidium bromide staining. Marker results were scored as 1 when the target band was detected and 0 when the target band was not detected. Inoculation of strains For rice blast assays, controlled inoculation was performed using the rice-pathogenic M. oryzae strain Guy11, which was obtained from the Institute of Plant Protection, Jiangsu Academy of Agricultural Sciences. Because Guy11 represents a single virulence background, the inoculation results were interpreted as resistance responses to this specific strain [36]. Leaf blast inoculation was conducted at the 3–4-leaf stage. Seedlings were acclimated in a greenhouse at 25–28 °C for 12 h before inoculation. A conidial suspension containing 0.8% (v/v) Tween 20 was sprayed onto the leaves until the leaf surface was evenly covered with droplets. After inoculation, plants were kept in darkness at 25–28 °C and ≥ 95% relative humidity for 24 h, and then transferred back to light conditions. High humidity was maintained by misting three to four times daily. Two biological replicates were included. Panicle blast inoculation was performed by injection. At the peak tillering stage, the conidial suspension was injected into the leaf sheath. Each plot consisted of one row with 12 plants. Three tillers per plant were inoculated, with approximately 1 mL of suspension injected into each culm. At the booting stage, the conidial suspension was injected into the culm 5–10 cm below the panicle apex until the suspension exuded from the emerging leaf. For BB assays, the inoculum consisted of a mixture of the dominant Hainan race HN01 of Xanthomonas oryzae pv. oryzae [37] and the Philippine differential race PXO99 (P6) [38]. The bacterial suspension was adjusted to approximately 1 × 10⁸ cells mL⁻¹. Inoculation was performed in the afternoon on sunny days at the peak tillering stage. For each plot, fully expanded leaves from five plants were selected and inoculated using the leaf-clipping method. Resistance Evaluation Rice blast resistance was evaluated under field conditions following artificial inoculation, with a primary focus on panicle blast. Disease assessment and integrated evaluation were conducted according to the Chinese agricultural standard NY/T 2646 − 2014. Leaf blast was scored when the susceptible control reached a disease grade of ≥ 7. For each accession, the three most severely diseased plants were selected, and the most severely infected leaf on each plant was scored. The mean score of these three plants was recorded as the leaf blast score, and the maximum leaf blast score was used for subsequent index calculation. Panicle blast was assessed at the early yellow-ripening stage, when approximately 80% of the upper spikelets had matured. For each accession, the most severely diseased panicles were evaluated to determine the panicle blast incidence score and panicle blast loss-rate score. The integrated blast resistance index (IBRI) was calculated as follows: IBRI = 0.25 × LBS + 0.25 × PBS + 0.50 × PBLS, where LBS, PBS, and PBLS represent the leaf blast score, panicle blast incidence score, and panicle blast loss-rate score, respectively (Table S1). BB was evaluated according to HNZ070-2015. Disease severity was recorded at 14 days after inoculation. For each accession, the five most severely diseased plants were selected, and the most severely infected leaf on each plant was assessed. The mean value of these five plants was used as the final value. Data analysis The number of marker-detected resistance-associated loci was summarized for each accession. Spearman rank correlation analyses were performed using SPSS 14.0 to evaluate associations between marker-detected resistance-associated loci and resistance phenotypes. These analyses were used to describe statistical associations rather than infer the functional effects of individual loci. Statistical significance was determined at P < 0.05. Results Distribution and frequency of rice blast resistance genes All seven maintainer lines showed at least two positive marker signals, with the number of detected loci ranging from 2 to 5 per line (Fig. S1; Table 2). Only five loci—Pik, Pib, Ptr, Pikh, and Pikm—were detected in this group. Pik and Ptr showed the highest detection frequencies, with positive signals in all seven maintainers (100%, 7/7), followed by Pib (85.7%, 6/7), Pikh (42.9%, 3/7), and Pikm (14.3%, 1/7). Huhan7B was positive for all five detected loci, representing the broadest marker-detected resistance-locus profile among the maintainer lines. Table 2Marker-based detection of target alleles at eleven rice blast resistance loci in 47 core WDR parental linesAccessionPikPibPtrPikhPikmPitaPi9Pi5PiaPi2PigmTotal detected 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(%)53.255.397.942.68.525.519.142.617.036.20.0——“1” indicates detection of the target resistance-associated allele or diagnostic haplotype defined by the corresponding marker, whereas “0” indicates no detection. “Total detected loci” denotes the number of blast resistance loci for which the target allele or haplotype was detected in each accession. Frequency (%) was calculated based on all 47 accessions Table 2Marker-based detection of target alleles at eleven rice blast resistance loci in 47 core WDR parental linesAccessionPikPibPtrPikhPikmPitaPi9Pi5PiaPi2PigmTotal detected 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(%)53.255.397.942.68.525.519.142.617.036.20.0——“1” indicates detection of the target resistance-associated allele or diagnostic haplotype defined by the corresponding marker, whereas “0” indicates no detection. “Total detected loci” denotes the number of blast resistance loci for which the target allele or haplotype was detected in each accession. Frequency (%) was calculated based on all 47 accessions Marker-based detection of target alleles at eleven rice blast resistance loci in 47 core WDR parental lines “1” indicates detection of the target resistance-associated allele or diagnostic haplotype defined by the corresponding marker, whereas “0” indicates no detection. “Total detected loci” denotes the number of blast resistance loci for which the target allele or haplotype was detected in each accession. Frequency (%) was calculated based on all 47 accessions Among the 26 restorer lines, Huhui15 showed no positive marker signal, whereas Hanhui3 was positive only for Ptr (Fig. S1; Table 2). Overall, 57.7% (15/26) of the restorers carried positive marker signals at three or more loci. Ptr was the most prevalent locus, detected in 96.2% (25/26) of the lines, followed by Pi5 (50.0%, 13/26), Pi2 (38.5%, 10/26), and Pi9 (34.6%, 9/26). Among the Pik-family loci, Pikh, Pik, and Pikm were detected in 30.8% (8/26), 23.1% (6/26), and 11.5% (3/26) of the restorers, respectively. Pib was also detected in 30.8% (8/26) of the lines. Pita and Pia were each detected in 19.2% (5/26) of the restorers, whereas no Pigm-associated positive signal was observed. Based on the number of detected loci, 3.8% (1/26) of the restorer lines showed no positive signal, 38.5% (10/26) were positive for 1–2 loci, 34.6% (9/26) for 3–5 loci, and 23.1% (6/26) for six or more loci. Hanhui68 showed the broadest detected profile, with positive signals for seven loci (Pik, Pib, Ptr, Pikh, Pita, Pi5, and Pi2), whereas Hanhui49, Hanhui56, Hanhui308, Hanhui625, and Hanhui808 were each positive for six loci. All 14 conventional rice accessions were positive for at least three detected loci, with individual counts ranging from 3 to 7 (Fig. S1; Table 2). Ptr was detected in all accessions (100%). Pik and Pib were also highly prevalent, each detected in 85.7% (12/14) of the accessions, followed by Pikh at 64.3% (9/14). Pita, Pi5, and Pi2 were each detected at a moderate frequency of 50.0% (7/14), whereas Pia showed a lower detection frequency of 21.4% (3/14). In contrast, Pikm, Pi9, and Pigm were not detected in this group. Huhan1509 showed the broadest marker-detected profile, with positive signals for seven loci, and five accessions—Huhan932, Huhan1171, Huhan1516, Huhan1517, and Huhan1573—were each positive for six loci (Table 2). All seven maintainer lines showed at least two positive marker signals, with 2–5 detected loci per line (Fig. S1; Table 2). Only five loci—Pik, Pib, Ptr, Pikh, and Pikm—were detected in this group. Pik and Ptr were detected in all seven maintainers (100%, 7/7), followed by Pib (85.7%, 6/7), Pikh (42.9%, 3/7), and Pikm (14.3%, 1/7). Huhan7B was positive for all five loci and showed the broadest marker-detected profile among the maintainer lines. Among the 26 restorer lines, Huhui15 showed no positive marker signal, whereas Hanhui3 was positive only for Ptr (Fig. S1; Table 2). Overall, 57.7% (15/26) of restorers were positive for three or more loci. Ptr was the most frequent locus (96.2%, 25/26), followed by Pi5 (50.0%, 13/26), Pi2 (38.5%, 10/26), and Pi9 (34.6%, 9/26). Pikh and Pib were each detected in 30.8% (8/26) of restorers, followed by Pik (23.1%, 6/26) and Pikm (11.5%, 3/26). Pita and Pia were each detected in 19.2% (5/26), whereas no Pigm-associated positive signal was observed. Based on the number of detected loci, 3.8% (1/26) of restorers showed no positive signal, 38.5% (10/26) were positive for 1–2 loci, 34.6% (9/26) for 3–5 loci, and 23.1% (6/26) for six or more loci. Hanhui68 showed the broadest profile, with positive signals for seven loci (Pik, Pib, Ptr, Pikh, Pita, Pi5, and Pi2), whereas Hanhui49, Hanhui56, Hanhui308, Hanhui625, and Hanhui808 were each positive for six loci. All 14 conventional rice accessions were positive for at least three loci, with 3–7 detected loci per accession (Fig. S1; Table 2). Ptr was detected in all accessions (100%). Pik and Pib were also frequent, each detected in 85.7% (12/14), followed by Pikh (64.3%, 9/14). Pita, Pi5, and Pi2 were each detected in 50.0% (7/14), whereas Pia was detected in 21.4% (3/14). Pikm, Pi9, and Pigm were not detected in this group. Huhan1509 showed the broadest marker-detected profile, with seven detected loci, and five accessions—Huhan932, Huhan1171, Huhan1516, Huhan1517, and Huhan1573—were each positive for six loci (Table 2). Distribution and frequency of BB resistance genes Using gene-specific markers, we genotyped 47 core WDR parental lines for four BB resistance loci: xa5, Xa23, Xa7, and Xa21 (Fig. S2; Table 3). Positive marker signals were unevenly distributed among parental groups, with 0–3 detected loci per accession. Overall, only 34.0% (16/47) of the accessions were positive for at least one BB resistance locus. The restorer group showed the highest detection rate (57.7%, 15/26). Within this group, Hanhui303 and Hanhui708 showed the broadest profiles, with positive signals for three loci (Xa23, Xa7, and Xa21), whereas Hanhui625 and Hanhui366 were each positive for two loci. In contrast, no positive signal for any of the four BB loci was detected in the seven maintainer lines. Among conventional accessions, only Yanhan6 was positive (1/14), while the remaining 13 accessions were negative for all tested loci. Table 3Marker-based detection of target alleles at four BB resistance loci in 47 core WDR parental linesAccessionxa5Xa23Xa7Xa21Total detected lociHuhan1B00000Huhan5B00000Huhan7B00000Huhan17B00000Huhan39B00000Huhan41B00000Huhan81B00000Huhan1500000Huhan1900000Huhan93200000Huhan116400000Huhan116600000Huhan117100000Huhan120100000Huhan150900000Huhan151200000Huhan151600000Huhan151700000Huhan152300000Huhan157300000Yanhan601102Hanhui300101Hanhui900000Hanhui1500000Hanhui2100000Hanhui3700000Hanhui4900000Hanhui5601001Hanhui5700101Hanhui6800000Hanhui30301113Hanhui30801001Hanhui36600112Hanhui50500011Hanhui50801102Hanhui60100000Hanhui62501102Hanhui70801113Hanhui78001001Hanhui80801001Hanhui84000101Hanhui98100000Hanhui301500000Hanhui301600000Hanhui164000000Hanhui820000101Hanhui820901102R gene frequency(%)021.323.48.5——“1” indicates detection of the target resistance-associated allele or diagnostic haplotype defined by the corresponding marker, whereas “0” indicates no detection. “Total detected loci” denotes the number of BB resistance loci for which the target allele or haplotype was detected in each accession. Frequency (%) was calculated based on all 47 accessions Table 3Marker-based detection of target alleles at four BB resistance loci in 47 core WDR parental linesAccessionxa5Xa23Xa7Xa21Total detected lociHuhan1B00000Huhan5B00000Huhan7B00000Huhan17B00000Huhan39B00000Huhan41B00000Huhan81B00000Huhan1500000Huhan1900000Huhan93200000Huhan116400000Huhan116600000Huhan117100000Huhan120100000Huhan150900000Huhan151200000Huhan151600000Huhan151700000Huhan152300000Huhan157300000Yanhan601102Hanhui300101Hanhui900000Hanhui1500000Hanhui2100000Hanhui3700000Hanhui4900000Hanhui5601001Hanhui5700101Hanhui6800000Hanhui30301113Hanhui30801001Hanhui36600112Hanhui50500011Hanhui50801102Hanhui60100000Hanhui62501102Hanhui70801113Hanhui78001001Hanhui80801001Hanhui84000101Hanhui98100000Hanhui301500000Hanhui301600000Hanhui164000000Hanhui820000101Hanhui820901102R gene frequency(%)021.323.48.5——“1” indicates detection of the target resistance-associated allele or diagnostic haplotype defined by the corresponding marker, whereas “0” indicates no detection. “Total detected loci” denotes the number of BB resistance loci for which the target allele or haplotype was detected in each accession. Frequency (%) was calculated based on all 47 accessions Marker-based detection of target alleles at four BB resistance loci in 47 core WDR parental lines “1” indicates detection of the target resistance-associated allele or diagnostic haplotype defined by the corresponding marker, whereas “0” indicates no detection. “Total detected loci” denotes the number of BB resistance loci for which the target allele or haplotype was detected in each accession. Frequency (%) was calculated based on all 47 accessions At the individual-locus level, Xa23 and Xa7 were the most frequently detected loci, with frequencies of 21.3% (10/47) and 23.4% (11/47), respectively. These two loci co-occurred in 12.8% (6/47) of the panel. Xa21 was detected at a lower frequency (8.5%, 4/47) and showed a scattered distribution. No xa5-associated positive marker signal was observed in any accession. Together, BB resistance-associated marker signals were mainly detected in the restorer group and were largely attributable to Xa23 and Xa7. The absence of positive signals in the maintainer lines suggests a major gap in BB resistance-locus deployment. The low frequency of Xa21 and the absence of xa5 further indicate that these loci have been underused in the current WDR parental panel. Uniform nursery evaluation for rice blast and BB Based on resistance scores, with lower scores indicating stronger phenotypic resistance, the 47 core WDR parental lines were classified as highly resistant (score ≤ 2), moderately resistant (score 3–5), or susceptible (score ≥ 6). Differences in the number of marker-detected resistance loci were observed among resistance categories for both diseases. For rice blast, highly resistant lines generally carried more resistance genes than susceptible lines. This inverse relationship was statistically supported by a significant negative correlation (Spearman’s ρ = -0.330, P = 0.023) (Fig. 1).In this study, higher disease scores indicate greater susceptibility, whereas lower scores indicate stronger resistance. Therefore, the decrease in blast scores with increasing numbers of detected resistance genes was reflected as a negative correlation. For instance, Huhan1509 and Hanhui68, each carrying seven blast resistance genes, exhibited low resistance scores, whereas Hanhui15, with no detected blast resistance gene, showed a high score (Fig. 2A; Table 2; Table S3). Fig. 1Blast phenotypes of 47 core WDR parental lines. A Panicle blast phenotypes. B Leaf blast phenotypes. Scale bar = 2 cm. Accessions #1–#47 correspond to the numbering order listed in Table S4 Blast phenotypes of 47 core WDR parental lines. A Panicle blast phenotypes. B Leaf blast phenotypes. Scale bar = 2 cm. Accessions #1–#47 correspond to the numbering order listed in Table S4 Fig. 2Phenotypicdistribution of rice blast and BB resistance scores among 47 core WDR parental lines. A Rice blast resistance scores. B BB resistance scores. Resistance reactions were evaluated on a 0–9 scale, with lower scores indicating stronger phenotypic resistance Phenotypicdistribution of rice blast and BB resistance scores among 47 core WDR parental lines. A Rice blast resistance scores. B BB resistance scores. Resistance reactions were evaluated on a 0–9 scale, with lower scores indicating stronger phenotypic resistance For BB, the association between gene number and resistance was even more pronounced. Highly resistant lines carried substantially more BB resistance genes, resulting in a strongly significant negative correlation (Spearman’s ρ = -0.771, P < 0.001). Lines harboring BB resistance genes (e.g., Xa21, Xa23) typically displayed low scores, whereas those without any detected BB gene were predominantly moderate or susceptible (Fig. 2B; Table 3; Table S2) (Fig. 3). Fig. 3BB phenotypes of 47 core WDR parental lines. IRBB5, IRBB7, CBB21, and CBB23 were used as resistant controls. Scale bar = 2 cm. Accessions #1–#47 correspond to the numbering order listed in Table S4 BB phenotypes of 47 core WDR parental lines. IRBB5, IRBB7, CBB21, and CBB23 were used as resistant controls. Scale bar = 2 cm. Accessions #1–#47 correspond to the numbering order listed in Table S4 Overall, resistance scores varied substantially across the panel for both pathogens (Fig. 2). Rice blast scores were broadly distributed, with a number of accessions showing strong resistance (low scores). In contrast, BB scores were generally higher and more clustered, indicating a comparatively weaker overall resistance background against this disease in the germplasm (Fig. 1). Discussion This study characterized the marker-detected deployment of major blast and BB resistance loci in 47 core WDR parental lines. The results revealed partial fixation of framework loci and clear gaps in key complementary loci. These patterns support the need for genotype-guided parental selection and optimized pyramiding in WDR breeding. Blast resistance deployment in core WDR parents: patterns, gaps, and breeding priorities We genotyped 47 core WDR parental lines for 11 major blast resistance loci using diagnostic markers and evaluated field resistance to M. oryzae. Detection frequencies ranged from 0% to 97.9%, showing marked heterogeneity in marker-detected locus composition. In total, 40.4% of the lines were positive for three or fewer loci, indicating room for improvement in many breeding backgrounds. However, field resistance was not determined solely by the number of detected loci. These results suggest that future improvement of blast resistance in WDR parents may benefit from a shift from simple gene-number comparisons to locus-level prioritization and the rational assembly of complementary loci. Similar surveys in diverse rice panels reported wide variation in resistance-gene frequencies and limited multilocus deployment [39–41]. Consistent with these findings, our results support genotype-guided prioritization rather than simple gene counting in WDR parental improvement. We therefore further assessed whether key loci had been co-deployed in core WDR parents. Chromosomal hotspots and completion strategy of blast resistance in WDR parents Blast resistance loci are unevenly distributed in the rice genome and often cluster in chromosomal hotspots, particularly on chromosomes 6, 11, and 12 [42]. On chromosome 6, Pi9, Pi2, and Pigm form a tightly linked multi-allelic cluster [43]. Pi9 is widely used as a broad-spectrum resistance source, and the PigmR–PigmS module can enhance resistance while reducing yield penalty [44]. The Pigm–Pi2 combination may further broaden resistance in some backgrounds [45]. However, these loci were underdeployed in the WDR parental panel (Pi9, 19.1%; Pi2, 36.2%; Pigm, 0%), suggesting incomplete deployment of several potentially important backbone loci. In addition, potential interaction between the Pi9/Pi2/Pigm cluster and Pita-mediated immune signaling has been proposed [7], supporting coordinated deployment of these loci. On chromosome 12, Pita is tightly linked to Ptr (~ 100 kb), and previous studies have suggested possible cooperative interactions between these loci [8]. In our panel, Ptr was nearly fixed (97.9%), whereas Pita was detected in only 25.5% of the lines. This imbalance suggests that Ptr-linked haplotypes may have been retained during WDR breeding without consistent co-selection of the target Pita allele. On chromosome 11, Pik, Pikh, Pikm, and Pia showed uneven deployment. Pik and Pikh were relatively frequent (53.2% and 42.6%), whereas Pikm (8.5%) and Pia (17.0%) remained uncommon. Pib, located on chromosome 2, was also frequently detected (55.3%) and may serve as an additional framework locus. Together, these results support a stepwise improvement strategy for WDR parents: retain common framework loci such as Ptr, Pik, Pib, and Pikh; prioritize completion of underdeployed backbone loci, especially Pi9, Pi2, and Pigm; and add locally matched complementary loci, such as Pita, Pi5, Pia, or Pikm, according to target disease pressure. BB resistance deployment in core WDR parents: patterns, gaps, and breeding priorities Beyond blast resistance, we also examined the deployment of major BB resistance loci in core WDR parents. The target alleles or diagnostic haplotypes of Xa23 and Xa7, both associated with strong resistance across growth stages [12], were detected in fewer than 30% of the lines (21.3% and 23.4%, respectively). This indicates underdeployment of these two loci in the current WDR parental panel. Xa7 is particularly valuable in hot and humid regions because high temperature can enhance Xa7-mediated resistance [46, 47], and this gene is effective against the predominant pathotype IV in South China [13]. Xa21 was rarely detected (8.5%), whereas no xa5-associated positive signal was observed, despite its reported value against bacterial leaf streak [14]. These results suggest that BB improvement in WDR may benefit from prioritizing Xa23 and Xa7, deploy Xa7 preferentially in hot–humid target regions, and introduce Xa21 where broader pyramiding is needed. xa5 may be more suitable as a donor locus for specific high-quality inbred backgrounds [48]. Because BB pathotypes change over time and across regions, single-gene resistance is often race-dependent [2, 49]. Reduced resistance of xa5, Xa7, and Xa23 to pathotype IX has been reported [49], supporting the need for gene pyramiding to broaden resistance spectra and improve durability [50]. However, pyramiding should consider gene compatibility. xa5 may antagonize executor-type genes, such as Xa23, Xa27, and Xa7, because the corresponding TAL effectors may fail to activate executor-gene expression in an xa5 background [51, 52]. Therefore, BB pyramiding in WDR parents should prioritize loci with distinct mechanisms and use xa5–executor combinations cautiously unless their compatibility has been validated [53]. Practical validation of resistance-guided parental improvement in WDR The identified deployment gaps and pyramiding principles are supported by practical WDR breeding cases. Hanhui3 carries Xa7 and Ptr. After Pi9 introgression, the improved line Hanhui8200 retained drought resistance and drought-avoidance ability, while its blast score decreased from 7 to 1. Using Huhan7A as the female parent, the derived hybrids Hanyou 8200 and Hanliangyou 8200 were approved and showed strong drought resistance and high blast resistance [54]. Further introgression of Pi9, Xa23, and Bph6 into Hanhui3 generated Hanhui303. Hybrids derived from Huhan7A × Hanhui303 retained the main agronomic advantages of Hanyou73 and showed improved resistance to blast, BB, and brown planthopper [55]. In this study, Huhan7B, the maintainer line of Huhan7A, was positive for Pik, Pib, Ptr, Pikh, and Pikm. Given the close nuclear background generally shared by CMS-maintainer pairs, Huhan7A may share a similar resistance-locus profile; however, direct validation is still required. Previous MAS-based improvement of key WDR parents, such as Huhan1516 and Hanhui303, further supports the practical value of genotype-guided blast resistance improvement in WDR breeding. Together, these cases support the feasibility of genotype-guided parental improvement and compatible pyramiding in WDR hybrid breeding. The marker-detected locus profiles generated here provide a practical basis for MAS, WDR 6 K chip integration, high-throughput genotyping, and future molecular design breeding for disease resistance in WDR [40, 56]. Evidence boundaries and future validation This study characterized the marker-detected resistance gene/allele architecture of core WDR parental lines, not the functional contribution of each detected locus. Further validation using multiple representative pathogen isolates and functional analyses will be necessary to clarify the resistance spectra and contributions of individual loci and locus combinations [57, 58]. Data analysis All the phenotype and genotype data were arranged by Microsoft Excel and SPSS 14.0 (http://www-01.ibm.com/software/analytics/spsswas used to perform spearman rank correlation analyses. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3. Supplementary Material 4. Supplementary Material 4. Supplementary Material 5. Supplementary Material 5. Supplementary Material 6. Supplementary Material 6. Supplementary Material 7. Supplementary Material 7. Supplementary Material 8. Supplementary Material 8. Supplementary Material 9. Supplementary Material 9.