Fine mapping of qBK1.2, a major QTL governing resistance to bakanae disease in rice

Bakanae disease caused by Fusarium fujikuroi is an emerging disease of rice causing losses in all rice-growing regions around the world. A BC2F2 population was developed by backcrossing the recurrent parent Pusa Basmati 1121 (PB1121) with the recombinant inbred line RIL28, which harbors a major quantitative trait locus (QTL) governing resistance to bakanae, qBK1.2. MassARRAY-based single-nucleotide polymorphism (SNP) assays targeting the genomic region of qBK1.2 helped in fine mapping the QTL to a region of 130 kb between the SNP markers rs3164311 and rs3295562 using 24 recombinants. In-silico mining of the fine-mapped region identified 11 putative candidate genes with functions related to defense. The expression analysis identified two significantly differentially expressed genes, that is, LOC_Os01g06750 and LOC_Os01g06870, between the susceptible genotype PB1121 and the resistant genotypes Pusa1342 and R-NIL4. Furthermore, the SNPs identified in LOC_Os01g06750 produced minor substitutions of amino acids with no major effect on the resistance-related functional motifs. However, LOC_Os01g06870 had 21 amino acid substitutions, which led to the creation of the leucine-rich repeat (LRR) domain in the resistant genotype Pusa1342, thereby making it a potential candidate underlying the major bakanae-resistant QTL qBK1.2. The markers used in the fine mapping program are of immense utility in marker-assisted breeding for bakanae resistance in rice.


Introduction
Bakanae disease is one of the emerging diseases of rice incited by Fusarium fujikuroi (Nirenberg, 1976) (teleomorph: Gibberella fujikuroi Sawada (Ito and Kimura, 1931), a filamentous hyphomycete fungus.The disease is prevalent in temperate and tropical rice-growing regions of the world (Amoah et al., 1995;Desjardins et al., 1997;Amatulli et al., 2010;Egȇrci et al., 2021).Reported earlier as a minor disease, bakanae has emerged during the last decade as a serious problem in Basmati-growing areas in India, causing yield losses of up to 70% as well as impairing grain quality (Bashyal, 2018).The symptoms of the disease include seedling elongation, seedling mortality, reduced tillers, and elongated non-productive tillers (Bashyal et al., 2016).The disease is primarily seed borne, and secondary spread is through spores on plant parts and soil [Sun, 1975;Anderson and Webster, 2005].Hot water treatment, and chemical and biological control measures are the major interventions to manage the disease; however, genetic resistance is considered effective (Hayasaka et al., 2001;Hossain et al., 2015;Lee et al., 2018;Sarwar et al., 2018), as it is economical, sustainable, and eco-friendly.
Since the reported QTLs have been proven effective only against specific isolates of F. fujikuroi prevalent in different regions, the wide utilization of the reported QTLs in breeding is limited.However, Lee et al. (2018) found that the pyramiding of the two QTLs qBK1 and qBK1 wd provided a better level of resistance as compared to individual QTL introgressed lines.Identification of novel QTLs and their deployment into elite cultivars is a proven strategy to broaden the spectrum of resistance and restrict the evolution of pathogens.Furthermore, validating the identified QTLs is of utmost importance prior to their utilization in breeding programs.One of the approaches for validation of a QTL is the development of QTL introgressed near-isogenic lines (QTL-NILs), which can also be used for fine mapping and identification of underlying candidate genes.Also, developing functional markers for the candidate gene bears significance in improving precision in marker-assisted breeding.
In a population generated from the cross between PB1121 and Pusa1342, a major QTL qBK1.2 was mapped on chromosome 1, which explained 24.74% phenotypic variance (Fiyaz et al., 2016), but remains to be fine mapped.This QTL is effective against the F. fujikuroi isolates prevalent in Basmati-growing regions of India.Therefore, this study reports the validation and fine mapping of qBK1.2 while developing tightly linked markers for their use in marker-assisted breeding.

Plant material and development of mapping population
RIL28, an individual from a recombinant inbred line (RIL) population (derived from the cross between PB1121 and Pusa1342) carrying the major QTL for bakanae, qBK1.2, as confirmed by QTL flanking simple sequence repeat (SSR) markers RM10153 and RM5336, was backcrossed with the susceptible parent PB1121, and a backcross population was generated.The generation advancement was performed by shuttling the material between two locations, i.e., Indian Agricultural Research Institute-New Delhi (IARI-New Delhi) and Rice Breeding and Genetics Research Centre, Tamil Nadu (RBGRC-Tamil Nadu).The F 1 , BC 1 F 1 , and BC 2 F 1 generations were subjected to foreground selection using the markers RM10153 and RM5336 flanking the QTL qBK1.2, and background selection was carried out using 28 SSR markers polymorphic between the parental lines (PB1121 and RIL28) to estimate the recurrent parent genome (RPG) recovery.Background recovery was calculated using the formula (1 − (n/N), where n is the number of marker loci heterozygous or homozygous for the recurrent parent allele and N is the total polymorphic markers between parental lines.Genotyping was carried out in the BC 2 F 2 generation, and phenotyping for bakanae resistance was carried out in the BC 2 F 2:3 generation.The scheme utilized for developing the BC 2 F 2:3 population is provided in Figure 1.

Phenotypic screening against bakanae disease
F. fujikuroi isolates were multiplied on sterile sorghum grain at the Division of Plant Pathology, ICAR-Indian Agricultural Research Institute, New Delhi.A 15-day-old culture medium was mashed in sterile water and filtered using a muslin cloth.A suspension culture with a conidial concentration of 1 × 10 6 was used for inoculation.To screen the rice genotypes, the seeds were soaked in a suspension culture of a virulent F. fujikuroi isolate F250 (National Center for Biotechnology Information (NCBI); gene bank accession number KM50526, collected from the northwestern part of India) in the test tubes for 48 hours and incubated at a temperature of 30°C.These seeds were then sown in pot trays (Figure 2) and kept in a growth chamber with a day/night temperature of 30°C/25°C and a day/night relative humidity of 60%/80%.Disease scoring on mortality percentage was carried out 12 days after inoculation and classified into different classes (Table 1) as described by Fiyaz et al. (2014).

Genotypic screening during population development
Fresh leaf samples were collected from 15-day-old seedlings and crushed using a DNA-lyser in Eppendorf tubes.DNA was isolated using the cetyl trimethylammonium bromide (CTAB) protocol described by Murray and Thompson (1980).For the development of the backcross population, foreground selection was carried out using the markers RM10153 and RM5336 (Supplementary Table 1) flanking the QTL qBK1.2 (Supplementary Figures S1 and S2).Out of 119 SSR markers polymorphic between PB1121 and Pusa1342, 28 markers were found polymorphic between PB1121 and RIL28 (Supplementary Table 2).Parental polymorphism and foreground and background selection were performed in a PCR volume of 10 ml.Each 10-ml reaction volume included 2 ml of template DNA (50 ng), 1 ml of forward primer (5 pm/ml), 1 ml of reverse primer (5 pm/ ml), and 3 ml of Taq DNA Polymerase RED 2× master mix (AMPLIQON A/S, Odense, Denmark) and 3 ml of nuclease-free water.PCRs were performed in 96-well PCR plates with a thermal seal in a thermal cycler (Applied Biosystems Veriti ™ , Foster City, CA, USA).Amplified PCR products were resolved on 3.5% (w/v) agarose gel stained with ethidium bromide.Gels were visualized using a UV-transilluminator gel documentation system (Bio-Rad Gel Doc XR+, Hercules, CA, USA).

SNP genotyping of the recombinants BC 2 F 2:3 plants
Since there were no polymorphic SSR markers available within the region spanning RM10153-RM5336, the single-nucleotide polymorphisms (SNPs) polymorphic between PB1121 and Pusa1342 located between the markers, i.e., RM10153 and RM5336, were identified using the re-sequence data available with us (data unpublished).To design the SNP assay, 150-bp upstream and downstream sequences of the SNP were retrieved from the NCBI database using Oryza sativa cv.Nipponbare as the reference genotype.Then, the SNP assays were designed in assay design suite (ADS) software from Agena ® Bioscience (4755 Eastgate Mall, San Diego, CA, USA) using the SNP group file option, and SNP genotyping was performed using MassARRAY ® , which is based on matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry.The steps involved in the genotyping were the amplification of the target region harboring

2.5
In silico search and validation for candidate genes within the fine-mapped region RNA isolation of the treatment and control was performed 6 days post-inoculation from PB1121, RIL28, and R-NIL4 in three replicates.Total RNA was extracted from shoots using a NucleoSpin RNA kit (Macherey-Nagel, Düren, Germany).Genes were searched within the fine-mapped region using the genome browser in the Rice Genome Annotation Project database (http:// rice.uga.edu/)(Supplementary Table 3).Primers for real-time PCR analysis of putative candidate genes were designed using the NCBI primer blast tool with a parameter product length of 80-150 bp and Tm of 59°C-61°C.The sequences of these primers are listed in Table 2. Rice elF4a (eukaryotic initiation factor-4a) was used as the reference gene for the normalization of expression data.The PCR mixture contained 2.5 ml cDNA (10 times diluted), 5 ml of 2× SYBR green PCR master mix (Bio-Rad, USA), and 100 nM of each genespecific primer in a final volume of 10 ml.Real-time PCR was performed for all putative candidate genes.Negative template control (NTC) was also performed for each primer pair.Realtime PCR was performed in a Bio-Rad real-time PCR machine (Bio-Rad, USA).All PCRs were performed under the following conditions: 10 min at 95°C and 40 cycles of 15 s at 95°C, 30 s at 60°C and melt curve with a single reaction cycle following conditions 95°C for 15 s, 60°C for 1 min, and dissociation at 95°C for 15 s.Three biological replicates were analyzed for each sample.The relative expression ratio was calculated using the 2 −DDCT method (Livak and Schmittgen, 2001).

In silico analysis of candidate genes
The sequences of genes, i.e., LOC_Os01g06750 and LOC_Os01g06870 of PB1121 and Pusa1324, were generated by incorporating the SNPs, which were identified through resequencing of PB1121 and Pusa1342 using Nipponbare as a reference genome (Supplementary File Alignment).These  sequences were then subjected to SPLIGN analysis to identify if the introns were present (Kapustin et al., 2008).The resulting coding sequences (CDSs) thus generated were then translated, and alignment was performed to identify amino acid substitutions between PB1121 and Pusa1342 (Madeira et al., 2022).Using the translated proteins, prosite analysis was carried out for the identification of differences in defense-related motifs (De Castro et al., 2006).

Results
Among the RILs generated from the cross between PB1121 and Pusa1342, RIL28 was chosen for the study because it carried qBK1.2 and showed a percent seedling mortality of 32% as compared to 95.33% in the susceptible genotype PB1121 under bakanae infection in the artificial screening conditions (Figures 2A, B).Furthermore, RIL28 exhibited 76.47% similarity to PB1121 based on the 119 SSR markers of which 28 were polymorphic (Figure 3).
The F 1 s from the cross made between PB1121 and RIL28 when subjected to the test of hybridity using the qBK1.2linked markers, RM10153 and RM5336, showed one true hybrid out of 12 plants tested.A true F 1 was backcrossed with PB1121, 35 BC 1 F 1 s were generated with qBK1.2, and RPG recovery ranged from 78.57% and 82.14%.Out of these 35 BC 1 F 1 s, an individual plant with RPG recovery of 82.15% was backcrossed with PB1121 to generate 90 BC 2 F 1 s.The RPG recovery among the BC 2 F 1 plants carrying qBK1.2 ranged from 85.71% to 96.42%.A BC 2 F 1 plant with 96.42% RPG recovery was selfed to generate a BC 2 F 2 population comprising 1,100 individuals.Genotyping with foreground markers followed by phenotyping using F. fujikuroi F250 isolate (Figure 4) resulted in the identification of 24 BC 2 F 2:3 recombinants falling in both resistant and susceptible classes.

Disease reaction among the backcrossderived lines
The backcross population, i.e., BC 2 F 2:3 families, were phenotyped for bakanae resistance and classified based on disease reaction, i.e., percent seedling mortality.The NILs on artificial inoculation and screening showed significant variation in percent seedling mortality, which ranged from highly susceptible to highly resistant reactions (Figure 5).
Based on the marker profile and bakanae screening data of RM10153 and RM5336, a total of 24 recombinants were identified.These recombinants can be categorized into three groups, viz., i) recombinant on the RM10153 marker side, ii) recombinant on the RM5336 side, and iii) recombinant on both sides.Four SNPs within the target genomic region were identified based on whole genome re-sequencing data of the parental lines PB1121 and Pusa1342.Genotyping of recombinants with the MassARRAY (MALDI-TOF mass spectrometry)-based SNP genotyping delimited qBK1.2 to a region of 130 kb between the SNPs rs3164311 and rs3295562 (Figure 6).

Expression profiling of putative candidate genes
In-silico search within the fine-mapped region using a rice genome annotation project database (http://rice.uga.edu/)led to the identification of 21 putative gene models, among which 11 were with annotated disease resistance-related functions (SlVe1, SlVe2, Ve1, cf-2, and other Verticillium wilt resistance genes) (Figure 7).
Expression of all the 11 putative candidate genes was profiled under mock and inoculated conditions in PB1121, Pusa1342, and Resistant NIL-4 (R-NIL4).Two genes, that is, LOC_Os01g06750 and LOC_Os01g06870, showed significant upregulation when treated with F250 isolate of F. fujikuroi (day/night temperature of 30°C/25°C ( ± 3)°C and relative humidity of 60%/8%0 ( ± 10%)) in the resistant genotypes Pusa1342 and R-NIL4, while there was significant downregulation in the susceptible genotype PB1121.G e n e s s u c h a s LO C_ O s 0 1 g 06 7 2 0 , LO C _ O s 0 1 g 06 7 3 0 , LOC_Os01g06760, and LOC_Os01g06876 had significant differential expression between the resistant genotype Pusa1342 and susceptible genotype PB1121, but not in R-NIL4.The expression of LOC_Os01g06790 was significantly upregulated in R-NIL4 while downregulated in both Pusa1342 and PB1121 (Figure 8).This indicated that genes LOC_Os01g06750 and LOC_Os01g06870 could be the potential candidates underlying the major QTL qBK1.2.Further analysis indicated that these genes are orthologous to the tomato Verticillium wilt resistance gene Vei.

In silico comparison of candidate genes and products
LOC_Os01g06750 was predicted to possess five exons (Figure 9A).Alignment of the sequences of PB1121 and Pusa1342  Fine-mapped region of qBK1.2.The QTL region is delimited to 130 kb between SNPs rs3164311 and rs3295562 using recombinants.The values on the right side indicate percent mortality upon inoculation.NIL, near-isogenic line; R-NIL, resistant NIL; S-NIL, susceptible NIL; QTL, quantitative trait locus; SNPs, single-nucleotide polymorphisms.
led to the identification of one, six, and four SNPs on exons 1, 4, and 5, respectively.Alignment of 494 amino acid translated sequence of PB1121 and Pusa1342 (Figure 9B) resulted in the identification of three amino acid substitutions: i) substitution of threonine with proline at 185th amino acid, ii) serein with glycine at 191st amino acid, and iii) leucine with proline at 246th amino acid position.Since these substitutions did not lead to any structural differences (Figure 9C), the role of this gene at protein level cannot be confirmed.
LOC_Os01g06870 was predicted to be an intron-less gene that codes for resistance protein SlVe1 precursor with two leucine-rich repeat (LRR) domains.DNA sequence alignment led to the identification of 30 SNPs between PB1121 and Pusa1342.Translated product alignment resulted in the identification of 21 amino acid substitutions.This led to the creation of an extra LRR domain in the resistant parent Pusa1342 as compared to the susceptible genotype PB1121 (Figure 10).Therefore, LOC_Os01g06870 may be considered as the potential candidate underlying qBK1.2conferring tolerance to bakanae disease.

Discussion
The severity of bakanae disease has been increasing over the years in the Indo-Gangetic plains of India wherein Basmati rice is one of the major crops.The majority of the Basmati rice varieties show susceptibility to bakanae.Although genetic resistance to bakanae is reported by several studies, the sources of resistance lie outside the Basmati gene pool.Naturally, Basmati rice has a narrow genetic base, and hence, the deployment of resistance genes is sought from non-Basmati sources.This necessitates a long breeding process to safeguard the specific Basmati traits along with the recruited trait.Recently, it was established that markerassisted backcross breeding can ease this cumbersome breeding process and accelerate varietal development with enormous success (Lee et al., 2018).
The QTL qBK1.2 identified on chromosome 1 lies over a span of 260 kb on the short arm flanked by the SSR markers, i.e., RM10153 and RM5336 (Fiyaz et al., 2016).In the current study, we used the approach of fine mapping qBK1.2 using QTL-NILs.NILs provide Relative expression (log fold change) of putative candidate genes underlying the fine-mapped region.
several advantages such as low background noise and genetic similarity to the sensitive parent.Because of these benefits, NILs are considered ideal for QTL validation, fine mapping, comparative genomics, marker development, gene expression studies, and even deployment as new cultivars.The BC 2 F 2 population generated in the current study had an RPG recovery of 97.91%.Moreover, the F 2:3 population provided an opportunity for evaluating resistance reactions under the artificial screening system.The variation observed among the NILs was quantitative and could be due to environmental influence.The environmental influence of disease response makes the management of phenotyping cumbersome, expensive, and time-consuming.
The SNP genotyping among the recombinants delimited qBK1.2 to a region of 130 kb between the SNPs rs3164311 and rs3295562.It was found that when this region was absent in the S-NIL1 and S-NIL2, the disease intensity reached above 94% (Figure 4).Since the other regions between the flanking markers were similar to the donor, the likelihood of the region being the candidate locus was very high, making us conclude that the QTL resided in this region.A similar strategy was used to fine map several QTLs in rice including qBK1 (Lee et al., 2019).The finemapped region comprised 11 annotated genes with putative disease resistance function.The expression pattern of a gene under mock and challenged inoculation conditions provides significant contrast in its involvement in response under infection (Matićet al., 2016; C h e n g e t a l ., 2 0 2 0 ) .G e n e s L O C _ O s 0 1 g 4 1 7 7 0 a n d LOC_Os01g41780 were annotated to encode leucine-rich repeat motifs, which were identified to negatively regulate bakanae disease resistance (Lee et al., 2019).Eight of 11 genes that were subjected to expression profiling between the parental lines PB1121 and Pusa1342 showed significant differential expression.The observed differences may be attributed to the response to infection dynamics and/or due to inherent differences in the genomes of these genotypes.Therefore, comparing the expression between the NILs could help identify the genes responding to the infection.Accordingly, the R-NIL4 carrying qBK1.2 in the genetic background of PB1121 was profiled along with the parental lines.The expression pattern ruled out the involvement of five loci, i.e., LOC_Os01g06720, LOC_Os01g06730, LOC_Os01g06760, LOC_Os01g06876, LOC_Os01g06900, and LOC_Os01g06790, in the manifestation of resistance to bakanae because they failed to show a differential response between qBK1.2 carriers (R-NIL4 and Pusa1342) and the non-carrier (PB1121).Two genes, that is, LOC_Os01g06750 and LOC_Os01g06870, have shown significant differential expression between susceptible (PB1121) and resistant individuals (Pusa1342 and R-NIL4).A comparative analysis of encoded protein using BLASTp showed that these genes are orthologues of Verticillium wilt (Verticillium dahliae) resistance gene of tomato, i.e., Ve1, which is rich in leucine-rich repeats.Ve1 gene of tomato was identified to provide resistance in tomato against race 1 strains of V. dahliae and Verticillium albo-atrum (De Jonge et al., 2012).It was also shown that Ve1 requires the cascade of signaling genes, such as Enhanced Disease Susceptibility 1 (EDS1), Non-race-specific Disease Resistance 1 (NDR1), NB-LRR Protein Required for HR-Associated Cell Death 1 (NRC1), ACIF, MAPK/ERK kinase 2 (MEK2), and SERK3/BAK1 (Fradin et al., 2009).The other gene, i.e., Ve2, has been shown to act antagonistically to Ve1 using RNAi, and it was found that Ve1 regulates the expression of defense-related genes by minimizing the effects of Ve2 on these genes (Nazar et al., 2018).
LOC_Os01g06750 is an interrupted gene with four introns, while LOC_Os01g06870 is an uninterrupted gene.The gene sequence variation among the genotypes determines the target phenotype.Although SNPs were identified in LOC_Os01g06750 between PB1121 and Pusa1342, the translated product revealed minor substitution of amino acids with no major effect on functional motifs, i.e., leucine-rich repeat having a role in the resistance-related functions (McHale et al., 2006).This indicates its function in providing tolerance to bakanae at the transcriptional or posttranscriptional level or by generating protein products through alternate splicing of the exons.However, the translated product of LOC_Os01g06870 had 21 amino acid substitutions, which led to the creation of the LRR domain in the resistant individual P usa1342.The differential expression of LOC_Os01g06870 and functional difference in gene product make it the potential candidate gene in providing tolerance against bakanae disease and can be conclusively proved with functional validation by generating transgenic events in the susceptible genotype.
We conclude that LOC_Os01g06870 is a potential putative candidate gene conferring tolerance to bakanae disease, underlying the major QTL qBK1.2 between the SNPs rs3164311 and rs3295562.These markers are resources of immense importance for a rice breeding program targeting to introgress qBK1.2 for developing bakanae resistance in rice.

FIGURE 1
FIGURE 1Shuttle breeding scheme for rapid development of backcross mapping population.
FIGURE 2 (A) Pure culture isolated from diseased plants.(B) Autoclaved sorghum seeds were inoculated with pure culture for mass multiplication.(C) The suspension was prepared.(D) Concentration of the suspension was checked under a light microscope.(E) Rice seeds were put in Eppendorf tubes and inoculated with suspension culture for 48 h.(F) Sowing and disease scoring was performed in 14 × 7 pot trays.
FIGURE 5 (A) Number of classes in different groups based on percent seedling mortality (PSM).(B) disease reaction of recombinants.

FIGURE 7
FIGURE 7Putative genes underlying the fine-mapped region.Locus IDs in red have a putative function in defense against disease, and IDs in black represent retroposons.

FIGURE 8
FIGURE 8 FIGURE 9 (A) Structure of gene LOC_Os01g06750; solid yellow boxes represent exons.(B) Alignment of proteins from gene LOC_Os01g06750 from PB1121 and Pusa1342.(C) Number of LRR motif in PB1121 and Pusa1342.LRR, leucine-rich repeat.
FIGURE 10 (A) Structure of gene LOC_Os01g06870 represents single exons.(B) Alignment of proteins from gene LOC_Os01g06870 from PB1121 and Pusa1342.(C) Reduced number of LRR domains in PB1121 as compared to Pusa1342.LRR, leucine-rich repeat.

TABLE 1
Phenotypic classification based on percent seedling mortality.

TABLE 2
List of primers for real-time PCR for validation of candidate genes.