Refinement of rice blast disease resistance QTLs and gene networks through meta-QTL analysis

Abstract

Rice blast disease is the most devastating disease constraining crop productivity. Vertical resistance to blast disease is widely studied despite its instability. Clusters of genes or QTLs conferring blast resistance that offer durable horizontal resistance are important in resistance breeding. In this study, we aimed to refine the reported QTLs and identify stable meta-QTLs (MQTLs) associated with rice blast resistance. A total of 435 QTLs were used to project 71 MQTLs across all the rice chromosomes. As many as 199 putative rice blast resistance genes were identified within 53 MQTL regions. The genes included 48 characterized resistance gene analogs and related proteins, such as NBS–LRR type, LRR receptor-like kinase, NB-ARC domain, pathogenesis-related TF/ERF domain, elicitor-induced defense and proteins involved in defense signaling. MQTL regions with clusters of RGA were also identified. Fifteen highly significant MQTLs included 29 candidate genes and genes characterized for blast resistance, such as Piz, Nbs-Pi9, pi55-1, pi55-2, Pi3/Pi5-1, Pi3/Pi5-2, Pikh, Pi54, Pik/Pikm/Pikp, Pb1 and Pb2. Furthermore, the candidate genes (42) were associated with differential expression (in silico) in compatible and incompatible reactions upon disease infection. Moreover, nearly half of the genes within the MQTL regions were orthologous to those in O. sativa indica, Z. mays and A. thaliana, which confirmed their significance. The peak markers within three significant MQTLs differentiated blast-resistant and susceptible lines and serve as potential surrogates for the selection of blast-resistant lines. These MQTLs are potential candidates for durable and broad-spectrum rice blast resistance and could be utilized in blast resistance breeding.

Introduction

Rice is one of the major staple crops, feeding more than 30% of the global population. Expected rice production is threatened by an increased incidence of biotic and abiotic stresses hampering food security. Rice blast disease caused by Pyricularia oryzae Cavara (teleomorph: Magnaporthe oryzae B. Couch) is a destructive fungal disease that causes losses of 10–100%¹. It causes infection at all stages of the crop cycle, with an incidence in all agro-ecologies in both temperate and tropical environments. It is typically designated as leaf blast (leaves), collar blast (collar), neck blast (culm, culm node, panicle neck nodes), or panicle blast (panicles) based on the rice tissue part showing infection and symptoms. However, it is also known to affect roots². Magnaporthe is an ascomycete fungus that causes airborne disease to spread through conidia and is more predominant under high humidity³. It is globally regarded as the most notorious disease due to its distribution across 85 countries⁴.

The enormous diversity among the pathogenic M. oryzae races has a negative impact on its control and management. Cultural, chemical, and biological strategies have been developed and deployed to control this disease. Biological control, such as the use of biocontrol agents^5,6 and deployment of resistant cultivars, is an eco-friendly approach. However, wide application, effectiveness, and durability can be achieved only through the identification of stable resistant sources and their utilization in crop breeding. Moreover, molecular markers have been used effectively to reduce screening and increase the efficiency of transfer of resistance in terms of pace and precision. Over a hundred blast resistance genes have been reported across all the rice chromosomes^1,7. Most of these treatments are race specific and have limited durability. However, blast resistance seldom shows quantitative inheritance encompassing a cluster of genes¹. Riveting on race-nonspecific and broad-spectrum genes helps achieve durable horizontal resistance⁸. In this context, the discovery and characterization of quantitative trait loci (QTLs) governing rice blast resistance are of paramount interest.

Several QTLs linked to blast resistance in rice have been reported from different environments using different mapping populations^9–12. In addition, most of them (except for a few QTLs with major effects) are restrained from deployment in marker assisted selection (MAS), speculating uncertainty. The uncertainty or instability of QTLs is associated with different genetic backgrounds and growing environments¹³. QTL effects are influenced by epistasis and modifier effects, both of which are more sensitive to genetic background¹⁴. Furthermore, biased QTL-environment correlations will restrict the use of a QTL across different environments. Limited recombination in biparental mapping populations causes several problems, such as low population size, low logarithm of odds (LOD) score, genotyping errors, and phenotyping in minimal environments, which reduce the effectiveness of QTLs in MAS¹⁵. Despite these limitations, QTL mapping is a modest way to understand the genetics underlying the trait of interest¹⁶. Hence, identifying a robust consensus genomic region using QTL information from independent studies has found relevance. With the availability of vast rice genomic resources and databases, cataloging and compiling QTL information on a consensus map is feasible. Meta-QTL analysis is instrumental in integrating QTL information from independent studies to identify and project robust QTLs on a consensus map^16–18. This approach aims to discover whether the QTL for a trait or related traits from different studies colocalize on the consensus map. Furthermore, this method identifies ‘real QTL’ with refined QTL positions and narrows its confidence interval (CI)¹⁹. A narrowed CI of a QTL decreases the number of genes predicted within the interval by more than that of the original QTL, thereby increasing the recovery of probable candidates for the trait¹⁷.

With this perspective, we framed the study by hypothesizing the presence of stable large-effect QTLs associated with rice blast resistance in the rice genome. Meta-QTL analysis was implemented to test this hypothesis with the following objectives: (a) projection of MQTLs conferring rice blast resistance considering the original QTL; (b) identification of peak (nearest) markers linked to MQTLs; and (c) mining of the candidate genes underlying the MQTL region and predicting their role in the rice blast resistance.

Results

Discovery of MQTLs for rice blast resistance

Information on a total of 737 QTLs associated with rice blast resistance was collected from 53 independent studies reported over the past two decades (Table 1). These studies included recombinant inbred line (RIL, 40%), backcross (BC, 28%), F₂ (23%), and doubled haploid (DH, 9%) mapping populations (Fig. 1d) with sizes ranging from 31 to 1125 (Fig. 1a), which were evaluated for rice blast resistance across 14 different countries (Fig. 1e). Blast-resistant QTLs were reported across all the rice chromosomes, with the maximum occurring on chromosome 11 with 71 initial QTLs, and a minimum of 18 initial QTLs reported on chromosome 10 (Fig. 2a). The majority (682) of the initial QTLs were significant with a LOD > 2.5, except for a few (55) (Fig. 1b). The initial QTLs included 345 major-effect QTLs with a PVE ≥ 10%, while the remaining were minor-effect QTLs (Fig. 1c). A consensus map was generated using the individual maps from each of the 53 experiments.

Table 1.

Details of the QTL mapping studies on rice blast resistance used for meta-QTL analysis.

Sl. No.	Parentage	Mapping population type	Population size	Experiment place	Experiment year	No. of QTL	References
1	Akhanaphou × Leimaphou	RIL	103	India	2017	2	12
2	Suweon_365 × Ki_313	F2	69	Korea	2000	2	60
3	Pongsuseribu2 × Mahsuri	F2	320	Malaysia	2011	28	61
4	Mahsuri × Pongsuseribu2	F2	300	Malaysia	2012	14	62
5	Mahsuri × Pongsuseribu2	F2	188	Malaysia	2013	13	63
6	IR64 × Azucena	DH	114	India	2000	12	64
7	IR64 × Azucena	DH	93	India	2007	13	65
8	Zhenshan × Minguhi_63	RIL	241	China	2003	9	66
9	Suweon365 × Chucheong	RIL	190	Korea	2008	8	67
10	Heikezijing × Suyunuo	RIL	162	China	2013	2	68
11	Bodao × Suyunuo	RIL	212	China	2014	2	26
12	Nipponbare × Owarihatamochi	F2	146	Japan	2001	4	69
13	Fanny × Ol5	RIL	120	Colombia	2006	59	70
14	Dongnong415 × Lijiangxintuanheigu	RIL	114	China	2016	10	71
15	Heikezijing × Suyunuo	RIL	162	China	2017	5	72
16	IR64 × Azucena	F2	125	India	2003	3	73
17	Minguhi63 × Zhenshan97	RIL	241	China	2008	4	74
18	Bodao × Suyunuo	BC	155	China	2014	52	75
19	Miyazakimochi × Bikei22	BC	158	Japan	2014	2	76
20	Lemont × Jasmine_85	RIL	227	USA	2010	9	77
21	CR071 × Jin23b	BC	239	China	2020	47	10
22	Qingguai3 × Jin23b	BC	237	China	2020	27	10
23	Lemont × Jasmine_85	RIL	227	USA	2015	6	78
24	IR64 × Azucena	DH	111	Thailand	2010	8	79
25	Zhenshan_97 × Minghui_63	RIL	241	China	2008	29	80
26	Lijiangxintuanheigu × Sa0169	F2	122	Taiwan	2022	2	81
27	Danteswari × Dagaddeshi	BC	122	India	2017	5	82]
28	Kahei × Owarihatamochi	BC	158	Japan	2001	2	83
29	Ingngopportinawon × Koshihikari	BC	124	Japan	2014	1	84
30	Nekken2 × Hokuriko193	F2	193	Japan	2017	3	85
31	Kdml_105 × Jhn	RIL	587	Thailand	2006	7	86
32	Way_rarem × Oryzica_lianos5	BC	123	Indonesia	2011	16	87
33	Oryza rufipogon × Mr219	BC	261	Malaysia	2012	9	88
34	Oryza rufipogon × Mr219	BC	31	Malaysia	2012	13	88
35	Oryza minuta × Junambyeo	BC	112	Bangladesh	2011	5	89
36	Ingr15002 × Bpt5204	BC	188	India	2020	7	9
37	Tarommahalli × Khazar	F2	192	Iran	2011	7	90
38	IR64 × Azucena	DH	105	France	2003	9	91
39	Urn12 × Koshihikari	BC	39	Japan	2006	2	92
40	Heikezijing × Suyunuo	RIL	166	China	2010	22	93
41	CT9993 × Kdml105	RIL	141	Thailand	2002	5	94
42	Laka × Kencanabali	RIL	126	Indonesia	2016	1	95
43	IR64 × Jhn	F2	282	Thailand	2010	4	96
44	Lemont × Teqing	RIL	280	USA	2002	9	97
45	Bala × Azucena	RIL	205	Bangladesh	2005	21	98
46	Gigantevercelli × Vialonenano	F2	114	Italy	2016	2	99
47	Tianjingyeshengdao × Co39	RIL	363	China	2012	22	100
48	Jaohomnin × Rd6	BC	98	Thailand	2010	2	101
49	Indica9024 × Japonicalh422	BC	194	USA	1995	47	102
50	Zyo8 × Jx17	DH	127	China	2004	124	103
51	93_11 × Nipponbare	RIL	259	China	2013	7	104
52	Norin29 × Chubu32	F2	149	Japan	2002	2	105
53	Sanhuangzhan2 × Lijiangxintuanheigu	RIL	1123	China	2011	11	105

Figure 1.

Frequency distribution of (a) population size of each of the mapping populations used, (b) LOD score of the original QTLs and (c) phenotypic variation (%) explained by the original QTL. Graphical representation of the proportions of independent studies (d) using different biparental mapping populations and (d) countries from where the study is reported.

Figure 2.

Depiction of the decrease in the number of original QTLs through MQTL analysis as represented in the chromosome-wise distribution (a) and (b) the number of original QTLs within MQTLs.

In our study, all the chromosomes were projected with > 10 initial QTLs; hence, we followed Veyriera's approach for MQTL identification. A model with the lowest Akaike information criterion (AIC), corrected AIC (AICc), AIC3, Bayesian information criterion (BIC) and average weight of evidence (AWE) criteria was chosen for MQTL projection. Among the 737 initial QTL, only 435 QTLs were used in the meta-QTL projection due to low LOD scores and/or low PVE% and a large CI¹³ (Fig. S1). As many as 71 MQTLs were distributed across all chromosomes, with each MQTL consisting of one to 19 initial QTLs projected (Fig. 2). Interestingly, the confidence interval of the MQTL region decreased to an extent of 0.04 Mb, while as many as 60 MQTLs showed a confidence interval ≤ 2 Mb. The significance of each MQTL was depicted by its weight, which ranged from 0.03 to 0.62. A total of 31 MQTLs were associated with weight > 0.15 (Table S1). Among these, the top fifteen MQTLs with a confidence interval ≤ 2 Mb and a weight > 0.15 were considered highly significant (Table 2).

Table 2.

List of the 15 significant MQTLs associated with rice blast resistance.

MQTL	Chr	Position	Peak/Nearest marker	Weight	No. of genes in the interval	No. of original QTLs	Interval (Mb)	Genes related to rice blast resistance and RGA
M-QTL3.1	3	47.32	RM14875	0.19	47	10	0.39	Os03g0328000
M-QTL3.2	3	56.92	RM3033	0.18	27	8	0.19	Os03g0364400 (OsBBS1)
M-QTL3.4	3	108.76	RM6053	0.40	124	19	1.12	Os03g0674700 (OsGRF9), Os03g0680800 (OsBIHD1)
M-QTL4.5	4	115.89	RM17377	0.38	53	11	0.43	Os04g0573200 (SOD), Os04g0578000 (ACS2)
M-QTL6.3	6	38.58	RM19274	0.16	68	5	1.20	Os06g0286500 (Nbs1-Pi9) Os06g0286700 (Piz)
M-QTL6.4	6	62.73	RM20058	0.19	46	7	0.83	–
M-QTL6.7	6	110.28	RM20593	0.27	6	11	0.04	–
M-QTL8.4	8	87.74	RM23224	0.35	18	10	0.19	Os08g0448000 (4CL5)
M-QTL8.5	8	102.28	RM23434	0.19	114	5	0.78	Os08g0508800 (Lox2:Os:1), Os08g0509600 (WFP), Os08g0511700 (pi55-1), Os08g0512200 (pi55-2)
M-QTL9.2	9	38.49	RM24030	0.16	26	3	0.26	Os09g0327575, Os09g0327600 (Pi3), Os09g0327800 (Pi3)
M-QTL9.3	9	54.16	RM24236	0.33	62	8	0.90	Os09g0396900 (OsVIT2)
M-QTL9.4	9	83.38	RM24721	0.25	74	7	0.48	Os09g0533600 (OsRLCK278)
M-QTL10.3	10	85.01	RM25833	0.22	143	4	0.92	Os10g0536700 (prx128), Os10g0537800, Os10g0541000 (OsMLO4), Os10g0542800 (OsBSK1-2), Os10g0548200 (OsBSR-K1), Os10g0548300 (OsPUB53), Os10g0548700 (RLCK306), Os10g0552400 (OsPUB67), Os10g0552600 (OsHyPRP18), Os10g0552700 (OsHyPRP19)
M-QTL11.6	11	74.56	RM26835	0.19	42	18	0.68	–
M-QTL12.7	12	98.6	RM28598	0.15	76	6	0.84	Os12g0593000

Significant values are in bold.

Mining candidate genes within the MQTL regions

Using the flanking positions of each of the 71 MQTL regions, putative candidate genes linked to rice blast disease resistance were mined. In total, 8565 genes were identified within the MQTL region on all chromosomes according to the IRGSP 1.0 O. sativa japonica group genome. MQTL2.3 encompassed a maximum of 416 genes, while MQTL6.7 consisted of only 6 genes (Fig. S2). MQTLs were mined for resistance (R) gene analogs (RGAs) based on gene descriptions to obtain 48 characterized RGA candidates for blast resistance in 21 different MQTLs (Table S2). The RGA candidates coding for NBS–LRR-type resistance proteins on chromosome 11 included the characterized blast resistance genes RGA4 (Pia) and RGA5 (Pia) in MQTL11.1, PB1 in MQTL11.7, Pikh and Pi54 in MQTL11.8 and Pik in MQTL11.9. Genes encoding proteins similar to NBS–LRR-type resistance proteins, such as rNBS41, YR48 Nbs1-Pi9, RGA3, RPP8 and DRP-RP13, were found in MQTL2.1, MQTL5.3, MQTL6.3, MQTL11.1, MQTL11.8 and MQTL11.9, respectively. The genes deciphering the LRR domain-containing proteins LRR-RLK (MQTL1.5), OsIRL2 (MQTL2.3), Os_F0767 (MQTL2.5), OsFbox227 (MQTL4.4), OsFbox316 (MQTL6.6), and OsFbox661 (MQTL12.5) were also found. The LRR receptor-like kinase genes Docs1 and OsRPK1 were found in MQTL2.1 and MQTL5.3, respectively. Other receptor-like kinases involved in disease resistance, such as RLCK306 and OsLSK1, were found on chromosomes 1 and 10. Similarly, genes encoding NB-ARC domain-containing proteins, such as Xa39, YR1, OsOSC12, PB1, RPP13-like protein 3 and YR2/6/9/10/21/23, were identified. Pathogenesis-related TF/ERF domain-containing genes (OsPR10a, OsERF#045, OsWR4, Sub1C, AP2/EREBP#127, and OsERF#057) were found on six MQTLs. Furthermore, elicitor-induced defense response-related genes such as OsRLCK278, OsERF#039, Rpp17, UBC5B, and TPC1 were found within the identified MQTL regions. The genes encoding blast-induced MAPK signaling (OsMPK20-4), jasmonic acid-mediated defense signaling (OsRab11), and membrane-trafficking proteins involved in blast resistance (OsVAMP714) were found in MQTL1.5, MQTL6.5, and MQTL10.1, respectively. The pathogen-resistance protein PBZ1, encoded by OsPR10a, was identified in MQTL12.6. However, 35 different MQTLs were associated with as many as 125 uncharacterized RGA (Table S3). These 125 uncharacterized RGA included proteins similar to NBS–LRR (NLR), encoded by nine genes (six MQTLs); LRR kinase and domain-containing proteins, encoded by 26 genes (16 MQTLs); proteins similar to LRR, encoded by 17 genes (nine MQTLs); NB-ARC domain-containing proteins, encoded by 37 genes (26 MQTLs); and proteins similar to those encoded by 11 genes (three MQTLs). They also included plant disease resistance protein domains and other important yet uncharacterized RGA. Furthermore, Os10g0163040 in MQTL 10.1 was associated with a protein similar to the blast resistance protein.

Race-specific and broad-spectrum clusters of resistance (R) and/or (RGAs have been reported on rice chromosomes 1, 2, 4, 6, 11 and 12^10,20–22. Similarly, a few genes appeared to be in pairs or clusters within the identified MQTLs. For example, five genes (Os11g0224900/RGA3, Os11g0225100/RGA4/Pia, Os11g0225300/RGA5/Pia, Os11g0211300, and Os11g0213800) were identified in MQTL11.1; three genes (Os11g0493700, Os11g0492900 and Os11g0494100) were identified in MQTL11.5; two genes (Os11g0639100/Pi54/PIKH and Os11g0633500/RPP8) were identified in MQTL11.8; and three genes (Os11g0689100/PIKM/PIK, Os11g0688832/PIKM/PIK and Os11g0676200) were identified in MQTL11.9 (Tables 2, S3). Furthermore, clusters (≥ 5 genes) of different classes of RGAs, including those encoding NLR, NB-ARC, disease resistance proteins, and LRR proteins, were also observed. A cluster of five RGAs was observed within each of the MQTL1.5, MQTL4.4, MQTL5.3 and MQTL10.1. Similarly, clusters of six genes were observed within MQTL8.3 and MQTL12.6, while seven and eight RGAs were clustered within MQTL2.3 and MQTL12.2, respectively. Furthermore, five clusters were identified within chromosome 11 within MQTL11.1 (14 RGA), MQTL11.5 (7), MQTL11.7 (13), MQTL11.8 (6) and MQTL11.9 (21).

Genes specific to rice blast resistance were identified through a custom search in the Rice Annotation Project database (RAP-DB). Based on the reported characterized gene names, gene descriptions, and trait ontologies associated with blast resistance and characterized RGAs, a total of 199 candidate genes within 53 MQTLs were identified (Fig. 3, Table S4). These genes were associated with different defense mechanisms, such as MAPK signaling (e.g., OsMPK20-4/MQTL1.5), JA-induced defense response (NH1/MQTL1.2), R genes (Nbs1-Pi9/Piz/MQTL6.3), their analogs (e.g., OsRFP/MQTL1.5), and elicitors, such as PAMP (e.g., OsDjA9/MQTL6.1; OsRLCK55/MQTL1.8). A total of 13 characterized genes for blast resistance on MQTLs located on chromosomes 6, 8, 9 and 11 were identified in our study (Table 3). Seven of these genes were associated with significant MQTLs (weight > 0.15): Os06g0286700/Piz and Os06g0286500/Nbs1-Pi9 in MQTL6.3; Os08g0511700/pi55-1 and Os08g0512200/pi55-2 in MQTL8.5; Os09g0327600/Pi3 (Pi5-1) and Os09g0327800/Pi3 (Pi5-2) in MQTL9.2; and Os11g0639100/Pi54/Pikh in MQTL11.8 for blast resistance. Moreover, six genes characterized for blast resistance were identified within MQTL11.1 (Os11g0225300/Pia/RGA4 and Os11g0225100/Pia/RGA5), MQTL11.7 (Os11g0598500/Pb1) and MQTL11.9 (Os11g0682600/Pb2, Os11g0689100/Pik and Os11g0688832/Pik). Among the 15 most significant MQTL intervals, a total of 887 genes were identified, including 29 candidates for rice blast resistance (Table 2). These top 15 most significant MQTL regions included six characterized genes (Piz, Nbs-Pi9, pi55-1, pi55-2, Pi3/Pi5-1 and Pi3/Pi5-2). Furthermore, they included other important genes, such as Os10g0548700/RLCK306, Os10g0552400/OsPUB67, and Os10g0548300/OsPUB53, in MQTL10.3; these genes are associated with PUB proteins with U-box domains (Table S4).

Figure 3.

Representation of MQTLs harboring genetic loci associated with blast resistance. Loci represented in red indicates characterized blast resistance genes, green indicates RGAs, blue indicates genes associated with blast resistance as per the gene description, and black indicates loci with trait ontology associated with blast resistance.

Table 3.

Details of meta-QTL regions associated with characterized blast resistance genes.

Meta-QTL	Weight	Peak marker	MQTL interval (Mb)	Gene stable ID	Gene name	Gene description and gene synonyms
M-QTL6.3	0.16	RM19801	1.21	Os06g0286700	Piz	Similar to Piz(t)
				Os06g0286500	Nbs1-Pi9	Similar to NBS–LRR disease resistance protein homolog
M-QTL8.5	0.19	RM23442	0.78	Os08g0511700	pi55-1	LRR protein; candidate gene for pi-55
				Os08g0512200	pi55-2	Heavy metal associated domain; candidate gene for pi-55
M-QTL9.2	0.16	RM24030	0.26	Os09g0327600	Pi3	Similar to Pi5-1
				Os09g0327800		Similar to Pi5-2
M-QTL11.1	0.11	RM26220	1.72	Os11g0225100	Pia/RGA4	Nucleotide binding site-leucine rich repeats (NBS–LRRs) protein; RGA5, Pia-1
				Os11g0225300	Pia/RGA5	NBS–LRR protein, RGA4, Pia-2
M-QTL11.7	0.06	RM27016	0.96	Os11g0598500	Pb1	Pb1-like protein; NB-ARC domain containing disease resistance protein; Pb1
M-QTL11.8	0.20	RM4112	1.25	Os11g0639100	Pikh/Pi54	NBS–LRR protein, Broad spectrum resistance against M. oryzae; Pi-kh, Pi54
M-QTL11.9	0.13	RM27285	1.40	Os11g0682600	Pb2	NB-ARC domain containing disease resistance protein, Panicle blast resistance
				Os11g0689100	Pik/Pikp/Pikm	Coiled-coil NBS–LRR (CC-NBS–LRR) disease resistance protein; Pikp-2, Pikm6-NP, Pik6-NP
				Os11g0688832		CC-NBS–LRR disease resistance protein; Pikp-1, Pikm5-NP, Pik5-NP

Significant values are in bold.

Ontology of genes within the MQTL region

Gene ontologies of all the genes within MQTL regions were retrieved using the Ensembl database. Approximately 40% of the genes were associated with molecular functions, 29% were associated with cellular components, and 31% were associated with biological processes. The GO annotations of some of the genes involved in the biological processes included “autophagy”, “calcium ion transmembrane transport”, “cell cycle”, “cell division”, “cellular response to stimulus”, “defense response”, and “defense response to fungus”, which may have a role in rice blast disease resistance” (Fig. 4a). Furthermore, 12% of the genes were assigned to the “tolerance and resistance’ trait class, which included 32% of the disease resistance-conferring genes (Fig. 4b).

Figure 4.

Different (a) Gene Ontology terms and (b) trait classes associated with genes within M-QTL regions.

The functional annotation of 199 candidate genes related to rice blast resistance revealed their significant involvement in 30 different functions (Table S5) with their fold enrichment, and an FDR < 0.05. These genes were predominantly associated with gene networks related to the response to stimulus, the response to stress and the defense response, protein modification, signal transduction, and the diterpenoid and cinnamic acid biosynthetic/metabolic pathways. The number of genes shared among the functional annotations was represented by the network and clustering of pathways (Fig. S3). The genome-wide fold enrichment of the candidate genes indicated three enriched regions, two on chromosome 11 (MQTL11.8 and 11.9) and one on chromosome 6 (MQTL6.3), which were associated with genes characterized for blast resistance (Fig. S4). Furthermore, we considered only the 29 rice blast candidate genes within the 15 most significant MQTLs for functional annotation. Genes within eight of these MQTLs were associated with 11 different pathways (Fig. 5). The genes within MQTL6.3 and MQTL9.2 were associated with the nucleotide-binding adaptor (NB-ARC) shared by APAF-1, R proteins, and CED-4 domain, while those within MQTL8.5 were involved in the heavy metal-associated domain superfamily. Furthermore, the Os09g0327600/Os09g0327800/Pi3 genes in MQTL9.2 and the Os10g0542800/OsBSK1-2 genes in MQTL10.3 were affiliated with the plant defense pathway (Table 4).

Figure 5.

Functional annotation of genes related to rice blast resistance within the top 15 MQTLs. (a) Representation of the number of genes involved in different functional annotations (domain/motif/pathway) with fold enrichment. (b) network of genes involved in the expression of a function (domain/motif/pathway).

Table 4.

Functional annotation associated with rice blast resistance genes within the top 15 MQTLs.

Enrichment FDR	nGenes*	Total genes†	Fold enrichment	Functional annotation (domain/motif/pathway)	Genes	M-QTL
0.005	3	69	53.480	U box domain	OS10T0548300-01 OsJ_32366 OsJ_32402	M-QTL10.3
0.005	3	57	64.739	Modified RING finger domain	OS10T0548300-01 OsJ_32366 OsJ_32402	M-QTL10.3
0.012	8	1988	4.950	Coiled coil	OsJ_12119 b29O05.5 OS09T0327600-01 OS09T0327800-00 OsJ_32362 OS10T0548300-01 OsJ_32366 OsJ_32402	M-QTL10.3, M-QTL9.2
0.012	4	304	16.185	Plant defense	b29O05.5 OS09T0327600-01 OS10T0541000-01 BSK1-2	M-QTL9.2, M-QTL10.3
0.012	4	335	14.687	NB-ARC domain	b29O05.5 OS06T0286700-00 OS09T0327600-01 OS09T0327800-00	M-QTL6.3, M-QTL9.2
0.013	7	1530	5.628	Ion binding	GRF9 CCS 4CL5 CM-LOX1 SPL14 OS09T0396900-01 BSK1-2	M-QTL3.4, M-QTL8.4, M-QTL10.3, M-QTL9.3
0.019	2	32	76.877	Hydrophobic seed protein domain	OS10T0552600-01 OsJ_32400	M-QTL10.3
0.039	6	1390	5.310	ATP-binding	OS03T0364400-01 b29O05.5 4CL5 OS09T0327600-01 OsJ_30123 BSK1-2	M-QTL3.2
0.042	2	62	39.679	Heavy metal-associated domain superfamily	CCS OS08T0512200-00	M-QTL8.5
0.042	3	259	14.248	Rx, N-terminal	b29O05.5 OS09T0327600-01 OS09T0327800-00	M-QTL9.2
0.042	6	1516	4.868	Nucleotide-binding	OS03T0364400-01 b29O05.5 4CL5 OS09T0327600-01 OsJ_30123 BSK1-2	M-QTL3.2, M-QTL8.4, M-QTL9.2, M-Q TL10.3

*number of query genes involved in the function.

^†The total number of genes in the gene network involved in functional annotation.

Validation of MQTLs

Five blast-resistant rice genotypes (mean score ≤ 4.5) and five susceptible genotypes (mean score ≥ 7.5) were genotyped using peak/nearest markers associated with each of the significant MQTLs (Table S1). Three of the peak markers, viz., RM17377, 40N23r and Pikh, in MQTL4.5, MQTL9.2 and MQTL11.8, respectively, differentiated resistant and susceptible groups (Figs. 6, S5). Among them, 40N23r and Pikh were reported to be gene based markers for the Pi5 and Pikh/Pi54 genes, respectively²³. Thus, validation of these peak markers confirmed the significance of the corresponding MQTLs in blast disease resistance.

Figure 6.

Validation of the peak markers of three significant MQTLs for their association with blast disease resistance. Polymorphism between the five-blast disease-resistant (R) and susceptible (S) contrasts for the peak markers (a) RM17377 (MQTL4.5), (b) 40N23r (MQTL9.2) and (c) Pikh (MQTL11.8).

In silico validation of the candidate genes associated with rice blast resistance was performed by screening for microarray-based expression studies. The expression profiles of 150 blast resistance candidates identified within different MQTLs in our study were retrieved from the RXP_3001 dataset (gene expression profile in whole leaves inoculated with M. oryzae) available in the RiceXPRO database (https://ricexpro.dna.affrc.go.jp/)²⁴. The expression patterns of candidate genes at 1, 2, 3, and 5 days post inoculation (dpi) with two M. oryzae strains (P91-15B and Kyu77-07A) in the differential Nipponbare entries harboring the blast resistance genes Pia and Pish were examined. Among the 199 candidate genes, the expression profiles of 73 were not available in the database. Within the expression profiles of 126 genes, 51 were upregulated and 33 were downregulated at 1, 2, 3, and 5 dpi with both compatible and incompatible reactions (Fig. S6). The gene Os02g0584700 in MQTL2.3 showed differential expression between compatible and incompatible reactions with both strains, while the gene Os06g0163000/OsPUB70 in MQTL6.2 discerned differential expression with the Kyu77-07A strain (Table S3; Fig. S6). However, 42 genes, including eleven characterized genes, exhibited differential expression at different dpi. The genes Os09g0327600 and Os06g0286700, which were infected with P91-15B and Kyu77-07A, respectively, were downregulated at 5 dpi in incompatible reactions. The expression of Os08g0511700 and Os09g0327600 in both of these strains was upregulated at 1 dpi and downregulated at 5 dpi, respectively. Furthermore, the expression of the Os09g0327800 gene was downregulated in the resistant combinations but upregulated in the susceptible combinations at 1 dpi with Kyu77-07A and at 2 dpi with the P91-15B strain (Fig. 7). A similar trend was observed for Os11g0225300/Pia-2, whose expression was downregulated in the resistant combinations but upregulated in the susceptible reactions at 1 dpi and 5 dpi with Kyu77-07A; the reverse was true for the P91-15B strain. Os11g0225100/Pia-1 was upregulated in the incompatible reaction but downregulated in the compatible reaction at 1 dpi and 2 dpi with the P91-15B strain. The gene Os11g0598500/Pb1 was downregulated in all reactions except for the resistance reaction with the P91-15B strain at 5 dpi. The genes Os11g0639100/Pi54/Pikh, Os11g0682600/Pb2 and Os11g0689100/Pik/Pikm/Pikp were downregulated in the resistant reaction but upregulated in the susceptible reaction in all four treatments with the Kyu77-07A strain. However, the expression of these genes was unaltered in response to treatment with the P91-15B strain (Fig. 7).

Figure 7.

Graphical representation of the fold change in the expression of characterized blast resistance genes within significant MQTLs at 1, 2, 3- and 5-days post inoculation (dpi) of leaves with two M. oryzae strains from RiceXPro. Pia/Pish × P91-15B and PISH × Kyu77-07A indicate incompatible (resistant) reactions, and Pish × P91-15B and ∆PISH × Kyu77-07A indicate compatible (susceptible) reactions. P91-15B and Kyu77-07A are M. oryzae strains, while Pia, Pish and ∆PISH represent Nipponbare (NB) genotypes harboring the corresponding genes (Pia and Pish).

Orthology of genes identified in the MQTL region with other plant species

The query genes (8318) within the MQTL region were identified from the data available from the IRGSP genome of the O. sativa japonica (OSJ) group. A total of 5595 OSJ genes exhibited differential orthologous relationships with O. sativa indica (OSI), Zea mays (ZM), and Arabidopsis thaliana (AT) (Fig. 8). Among these, 3391 genes showed high confidence in orthologity (= 1) with one or more of the three genomes. Furthermore, nearly half of the query genes were orthologous to all three genomes (AT, OSI, and ZM). As expected, most of the OSJ genes were orthologous to OSI (5158), followed by ZM (3963) and AT (3086) (Fig. 8a). Among the 199-rice blast resistance candidate genes in the OSJ genome, AT, ZM, and OSI had 106, 141, and 168 orthologs, respectively. The orthologity of characterized blast resistance genes found in the intervals MQTL6.3, MQTL8.5, MQTL9.2, MQTL11.1, MQTL11.7, MQTL11.8 and MQTL11.9 was analyzed using the g:Profiler online tool. Among the thirteen characterized genes, only Os11g0688832/Pik/Pikp/Pikm was orthologous to the AT, ZM and OSI genes. Two genes (Os08g0512200/pi55-2 and Os09g0327600/Pi3) were orthologous to the Zm and OSI genes, while six genes (Os06g0286700/Piz, Os06g0286500/Nbs1-Pi9, Os08g0511700/pi55-1, Os11g0225100/Pia/RGA4, Os11g0639100/Pi54/Pikh and Os11g0689100/Pik/Pikp/Pikm) were orthologous to the OSI genes only. Furthermore, the gene Os08g0512200/pi55-2 showed high orthology with both ZM and OSI (Fig. 8).

Figure 8.

Number of O. sativa japonica group genes within the MQTL region showing (a) orthologous relationships (b) high orthology (= 1) with OSI- O. sativa indica, ZM- Zea mays and AT- Arabidopsis thaliana. (c) Circular diagram representing the orthology of identified genes (within the M-QTL region) with other plant species. The outer circle represents the chromosomes of OSJ (black), OSI (blue), ZM (green) and AT thaliana (red) under comparison. (d) Characterized genes showing orthology with related species.

Discussion

Sustained rice production to meet global food security needs has gained paramount importance with the increasing population. Rice productivity is significantly constrained by pests and diseases. These findings have led to the convergence of rice researchers’ focus on resistance breeding against these constraints. Among several diseases, blast disease is one of the most devastating diseases in rice, causing yield losses of up to 100%. More than 100 genes and 350 QTLs conferring blast resistance have been identified, 37 of which have been cloned and characterized^1,9,10. Most of these genes are race specific and impart highly fragile vertical resistance against blast disease. Diverse rice germplasms, including cultivars, landraces and wild species, are being screened for genes contributing to broad-spectrum resistance²⁵. Disease resistance conferred by one or two major genes is effective but prone to breakdown with the emergence of new strains. Hence, gene clusters that provide broad-spectrum resistance and thereby impart horizontal durable resistance should be focused on. QTLs linked to blast resistance are likely to harness horizontal resistance, which is durable¹⁵. However, the effectiveness of QTLs in MAS is limited by several factors, such as linkage drag due to the large QTL interval, the QTL projection model, and limited recombination events in the biparental mapping population¹⁵. Moreover, the instability of QTLs identified in a particular mapping population prevents their deployment in MAS across genetic backgrounds. Fine-mapping the QTLs and delimiting them to one or two genes, such as Pb-bd1 or Pi-69(t), followed by their characterization and utilization in breeding programme is effective^25,26. However, fine mapping each of the many QTLs is unrealistic because it is time- and resource-demanding⁹. In this regard, the meta-QTL analysis approach could be useful for refining QTL intervals and subsequently validating their association with trait of interest¹⁸.

In the present study, initial QTLs were available on all the chromosomes, depicting the quantitative genetic architecture of blast disease resistance. A total of 71 MQTLs were projected from 435 initial QTLs reported in 53 independent studies. A reduction in the number of MQTLs from the initial QTLs indicated the colocation of the QTL and shrinkage of their confidence interval. A similar pattern was reported for rice blast²⁷ and rice grain weight¹⁵. Furthermore, approximately 3390 genes within the MQTL region exhibited high orthologity with other subspecies of rice (O. sativa indica group), cereal species (Z. mays) and A. thaliana. The orthology of these genes with related species verified the significance of the respective MQTLs and reiterated their stability across species²⁸. The genes underlying the identified MQTLs were affiliated with ontologies such as response to stimulus, defense response, and response to fungus, which indicated their possible role in resistance against rice blast. Resistance to blast disease is conferred by resistance (R) genes, such as nucleotide binding site-leucine rich repeats (NBS–LRR or NLR), nucleotide binding adaptors shared by APAF-1, R proteins, and CED-4 (NB-ARC), disease resistance proteins (DRPs), leucine rich repeats (LRRs) and pathogenesis related proteins (PRs)²⁰. The identified MQTL regions consisted of 280 disease resistance genes. Among them, 199 genes (R genes and their analogs) within 29 MQTL regions were candidates for blast resistance involved in PAMP-triggered immunity, MAPK signaling and phytohormone signaling. As many as thirteen genes characterized for blast resistance, such as Piz, Nbs1-Pi9, pi55-1, pi55-2, Pi3, Pia/RGA4, Pia/RGA5, Pb1, Pi54, Pikh, Pb2 and Pik/Pikp/Pikm^29–36, resided within the identified MQTL regions. These MQTLs harboring one or more characterized genes could potentially contribute to blast resistance across different genetic backgrounds.

According to the gene ontology, blast resistance candidates within 29 MQTL regions were involved in different defense pathways. For instance, MQTL2.3 consists of the Os02g0571100/OsCPS2 and Os02g0570400/OsKS7 genes, which are affiliated with diterpenoid biosynthetic/metabolic pathways that are known to influence blast resistance in rice³⁷. Similarly, genes related to the defense response pathway were found within MQTL1.5, 2.3, 4.1, 9.2, 10.3, 11.1, 11.9 and 12.6, which signifies their role in plant defense. Furthermore, thirteen of the MQTLs on chromosomes 1, 2, 4, 5, 8, 10, 11 and 12 consisted of clusters of > 5 RGAs, and MQTL6.3 was enriched with a defense-related gene cluster on chromosome 6. These MQTL regions harboring gene clusters for blast resistance indicate their possible role in broad-spectrum resistance^10,21,22. Furthermore, gene clusters or pairs of NLRs found on chromosome 11 (MQTL11.1, 11.5, 11.8 and 11.9) have been proven to play a role in broad-spectrum resistance²⁰. Introgression of MQTLs with RGA clusters works akin to pyramiding genes in conferring broad-spectrum resistance; thus, introgression is advantageous over that of a single gene^20,21,38.

The identified highly significant (weight > 15) and high-resolution MQTLs encompassed 887 genes, including 29 rice blast-associated genes. Among the significant MQTLs, MQTL3.1 harbors the gene Os03g0324600, which encodes an NB-ARC domain-containing protein, while Os03g0328000 encodes a DOCK family guanine nucleotide exchange factor involved in the defense response against fungus³⁹. Similarly, the genes Os03g0674700/OsGRF9 and Os04g0578000/ACS2 in MQTL3.4 and MQTL4.5, respectively, are known to enhance resistance against blast disease^40,41. The characterized genes Os06g0286700/Piz and Os06g0286500/Nbs1-Pi9 in MQTL6.3 and Os09g0327600/Os09g0327800/Pi3 in MQTL9.2 encode RGA, which is known to confer broad-spectrum resistance against rice blast^10,42,43. Furthermore, the genes Os08g0511700/pi55-1 and Os08g0512200/pi55-2 in MQTL8.5 encode LRR and heavy metal-associated domain-containing proteins, which are candidates for pi55(t)³¹. Furthermore, the Os10g0548300/OsPUB53 gene in MQTL10.3 harbors U-box and modified ring finger domains, indicating its role in the plant defense pathway^44,45. MQTL11.8 harbors Pi54/Pikh, a characterized gene for broad-spectrum resistance against blast⁴⁶. These MQTLs span ≤ 2 Mb intervals and are amenable to marker-assisted selection (MAS)⁴⁷. The peak/nearest markers of significant MQTL4.5, MQTL9.2 and MQTL11.8 were validated for their association with blast disease resistance. Since the ten genotypes utilized for validation were chosen from a diverse set of genotypes, the results can extrapolate to any new set of rice genotypes expecting the similar results. However, one should be careful while extrapolating the findings as that these markers can be assayed only for resistance or susceptibility regulated by corresponding MQTL regions and not for complete resistance to blast disease which in this case is a quantitative trait known to be regulated by several loci. The in-silico validation results may be extrapolated to other genotypes based on the presence of corresponding MQTL. Apart from that, the information of MQTLs identified and validated in the present study can be utilized in genomic prediction-based trait improvement programs by considering MQTL information as additive effects while predicting using genomic information. These markers could be used as surrogates in MAS for blast resistance breeding. Furthermore, in silico expression analysis of Os11g0639100/Pi54/Pikh, which resides within MQTL11.8, revealed differential expression of the Kyu77-07A strain incompatible and incompatible reactions to leaf infection, which validated the significance of the MQTL. Similarly, the genes within MQTL2.3 (Os02g0584700), MQTL6.2 (Os06g0163000/OsPUB70) and MQTL11.9 (Os11g0682600/Pb2 and Os11g0689100/Pik) also exhibited similar expression patterns. Furthermore, the validation using peak markers suggests that the markers can be directly utilized to screen germplasm or new breeding lines for identification of resistant lines. Since the ten genotypes utilized for validation were chosen from diverse set of genotypes, the results can extrapolate to any set of genotypes expecting the similar results. However, one should be careful while extrapolating that these markers can be assayed only for resistance or susceptibility regulated by corresponding MQTL regions and not for complete resistance to blast disease which a quantitative trait known to regulated by several loci. On the other hand, the results of the in-silico validation of genes in significant MQTL regions helps to understand the genes regulate resistance and susceptibility. The in-silico validation results may be extrapolated to other genotypes based on the presence of corresponding MQTL. The genes on chromosome 9 and 11 correspond to MQTL 9.2 and MQTL 11.8 were showed to differential expressions in in silico validation. These two MQTL were validated using peak markers for differentiating susceptible and resistant genotypes. Hence, results are validated to have significant scope to utilize in improvement of resistance in rice. Apart from that, the information of MQTLs identified and genes validated in the present study can be utilized in genomic prediction-based trait improvement programs by considering MQTL information as additive effects while predicting using genomic information⁴⁷. The results presented in our study will help researchers decipher the complex network of genes and pathways underlying rice blast resistance. Furthermore, the identified MQTLs have potential utility in blast disease resistance breeding programs.

Materials and methods

Literature survey and data collation

An exhaustive bibliographic survey related to QTL mapping for blast resistance in rice has been performed (https://scholar.google.com/, https://www.researchgate.net/) using keywords like rice blast, QTL, mapping, rice blast resistance, blast resistant genomic regions, blast QTL information, blast resistance genes etc. to retrieve the literature. Information on the size, parentage, and type of mapping population, mapping function, LOD score, and PVE (%) was collected for a total of 737 QTLs compiled from 53 studies reported from 14 different countries across the globe (Table 1). The mapping studies included F₂, doubled haploids (DH), recombinant inbred lines (RIL), and backcross (BC) mapping populations, with population sizes ranging from 31 to 1123. Input files, i.e., genetic map information and QTL information, were prepared as per the requirements of Biomercator V4.2^17,48. Those studies that lacked genetic map information were excluded from the study. The map information included the details of the type of mapping population and the genetic position of the markers, while the QTL information included the QTL ID, trait, trait ontology ID, experimental year and location, LOD score, PVE (%) value with the QTL position, and confidence interval, as described in the Biomercator V4.2 user guide.

Consensus map and QTL projection

The consensus map required for QTL projection was developed using Biomercator V4.2 by combining the genetic maps of the published original QTL studies and the reference map of rice available in the Gramene database (https://archive.gramene.org/db/markers/marker_view). Individual map files consisting of map and QTL information for each of the 737 original QTLs were integrated with the consensus map for their projection on the consensus map. Among them, 435 were projected on the consensus map for MQTL projection, while the other initial QTLs were not projected due to low LOD or PVE% and large CI¹³. For the projection of the SNP markers linked to the original QTL, the nearest SSR marker on the reference map corresponding to the position of the SNP was considered¹⁵. The QTL position and confidence interval of the original QTL were used where available; otherwise, the CI was calculated using the Darvasi and Soller equation⁴⁹, $\text{CI}=\frac{530}{\text{N}\ast {\text{R}}^2}$ for the F₂ and BC populations and their modifications, $\text{CI}=\frac{287}{\text{N}\ast {\text{R}}^2}$ and $\text{CI}=\frac{163}{\text{N}\ast {\text{R}}^2}$ for the DH and RIL populations, respectively, in which N is the size of the population and R² is the phenotypic variation explained by the QTL⁵⁰.

Meta-QTL analysis

Meta-QTL analysis was implemented on the consensus map of each chromosome with projected QTLs considering default parameters in Biomercator V4.2. As the initial QTL number was more than nine on each chromosome, the two-step ‘Veyrieras’ method of meta-analysis was used¹⁶. The Akaike information criterion (AIC), corrected AIC (AICc), AIC3, Bayesian information criterion (BIC) and average weight of evidence (AWE) were used to choose the best-fit model. The model with the lowest AIC value is the best-fit model, as it indicates the least loss of information²⁷. A best-fit model was chosen, and the other parameters were retained as defaults to obtain meta-QTLs. Based on the information from the best-fit model, the significant MQTLs are listed along with the complete information. The MQTL weights generated by the software were used to sort the identified MQTLs into significant MQTL.

Validation of MQTLs

The peak or nearest marker to the position of the significant MQTL was considered for validation (Table S1). The sequences of these primers were retrieved from the Gramene Marker database (https://archive.gramene.org/markers/) and synthesized for use in marker assays. Ten genotypes with differential responses to blast disease were selected from one of our experiments at the ICAR-NRRI and evaluated over two years (unpublished; Table S6). Disease scoring was performed following a standard protocol, and entries with a mean disease score ≤ 5.5 were considered to indicate quantitative resistance⁵¹. For the validation of peak markers of the significant MQTLs, we considered five entries with a mean disease score ≤ 4.5 as resistant and five entries with a mean disease score ≥ 7.5 as susceptible. DNA was extracted from leaf samples of ten genotypes using the CTAB method⁵². Genomic DNA was subjected to thermal amplification using corresponding primers (peak markers), and the amplified product was visualized through gel electrophoresis and a documentation system. The markers that differentiated the resistant and susceptible entries validated the corresponding MQTLs for blast resistance.

Mining candidate genes within the MQTL region

The physical position of the flanking markers for each MQTL on the consensus map was identified using a reference map. Using the ‘Biomart’ tool of Ensembl Plants⁵³ and the Gramene database (https://archive.gramene.org/plant_ontology/index.html), reported genes within the MQTL region were downloaded using the IRGSP 1.0 O. sativa japonica genome as a reference. The following gene sequences were used to identify RGA genes: NBS–LRR, LRR containing protein or kinase, MAPK, pathogenesis-related, NB-ARC domain containing genes, and elicitor protein. The gene description, gene ontology, and trait ontology related to blast disease resistance were identified using the RAP-DB (https://rapdb.dna.affrc.go.jp/search/) and Oryzabase databases (https://shigen.nig.ac.jp/rice/oryzabase/download/gene). Functional annotation (gene ontology (GO) and relationships between query genes based on GO terms, gene functions, etc.) was performed using the shinyGO v0.741 (http://bioinformatics.sdstate.edu/go74/) online tool⁵⁴. The Rice Expression Profile Database (RiceXPRO), a repository of 40 K microarray-based expression profiles, was utilized in our study because it is based on gene models in the RAP-DB. Furthermore, the RXP_3001 dataset was chosen for analysis. This dataset included the expression profiles of the genes in the whole leaf of Oryza sativa cv. Nipponbare inoculated with M. oryzae strains P91-15B and Kyu-077A in relation to water (H₂O) treatment. In our study, the expression profiles of 199 candidate genes identified within 29 MQTLs were extracted from the RXP_3001 dataset in the RiceXPro Version 3 online tool/database (https://ricexpro.dna.affrc.go.jp/).

Phenotyping of rice lines for blast disease response

Screening for rice blast disease response was carried out in a standard uniform blast nursery (UBN) described previously⁵⁵. Briefly, each test entry (20–30 plants/test entry) was raised in 1 m long rows on raised nursery beds with a 10 cm row spacing. One row of HR-12 (susceptible check) was sown after every 5 test entries and also along the boundaries to ensure adequate inoculum build-up and disease spread. Entries were screened in a UBN using augmented design during Kharif 2022 and 2023. In our experiment, we used a highly virulent strain MoK19-18 (NCBI GenBank Accession No. MT757287) isolated from a popular mega rice cultivar BPT-5204 following the spore drop technique⁵⁶. The mycelium and conidia were brushed into distilled water and filtered through muslin cloth. The spore concentration was adjusted approximately to 1 × 10⁵/mL⁻¹. The spore suspension containing Tween-20 (0.2%) was sprayed uniformly over all the entries after 25 days post-sowing (2–3 leaf stage) during the late evening hours, and plants were covered with a plastic tarpaulin overnight to create humid condition (> 90%). Further, water was sprayed 4–6 times a day for seven days to maintain high humidity and facilitate the disease development. Parallelly, blast disease pressure in the screening plots was augmented as described previously⁵⁷. Briefly, blast disease-infected leaf samples were collected from naturally infected diseased plants under field conditions and were incubated overnight in a polythene cover to promote sporulation. Later, M. oryzae spores were harvested, and the spore suspension was sprayed uniformly on all entries. This process was repeated continuously for 3–4 days in an interval of 3 days. This approach specifically helpful for slow blast phenotyping, where the disease occurrence is slow⁵⁷. Disease scoring was done after the susceptible check reached the highest score (9), approximately 7 days post inoculation, and scoring was repeated six times in 5-day intervals to avoid the escape of slow-blasting genotypes. A 0–9 scale devised by IRRI, Philippines⁵⁸, was used to record the disease reaction.

Orthology of genes in O. sativa indica, Z. mays and A. thaliana

The O. sativa indica, maize, and Arabidopsis orthologs of genes within the MQTL region were identified using the ‘Biomart’ tool of the Ensembl Plants database⁵³. The genes within MQTLs that were orthologous to the genes in the O. sativa indica, maize, and Arabidopsis genomes were visualized in a Circos plot using R software. The orthology of genes related to rice blast resistance was obtained through the g:Profiler online tool⁵⁹.

Author contributions

B.N.D., T.R.S.: concept; B.N.D., S.S.: collected the data; B.N.D., N.C.S., C.A., P.K.S., S.S.: analyzed the data; S.S., B.N.D., H.R., N.C.S., P.R.: prepared the tables; D.P.: blast disease screening of selected rice genotypes; N.C.S., B.N.D., C.A.: wrote the MS; T.R.S., S.S., L.B., J.L.K., C.P.: edited the MS.

Data availability

Data is provided within the manuscript and its supplementary information files.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-64142-0.