2022 Volume 97 Issue 5 Pages 229-235
Blast disease caused by the filamentous fungus Pyricularia oryzae (syn. Magnaporthe oryzae) is one of the most destructive diseases of rice (Oryza sativa L.) around the globe. An aus cultivar, Shoni, showed resistance against at least four Japanese P. oryzae isolates. To understand Shoni’s resistance against the P. oryzae isolate Naga69-150, genetic analysis was carried out using recombinant inbred lines developed by a cross between Shoni and the japonica cultivar Hitomebore, which is susceptible to Naga69-150. The result indicated that the resistance was controlled by a single locus, which was named Pi-Shoni. A QTL analysis identified Pi-Shoni as being located in the telomeric region of chromosome 11. A candidate gene approach in the region indicated that Pi-Shoni corresponds to the previously cloned Pik locus, and we named this allele Pikps. Loss of gene function mediated by RNA interference demonstrated that a head-to-head-orientated pair of NBS-LRR receptor genes (Pikps-1 and Pikps-2) are required for the Pikps-mediated resistance. Amino acid sequence comparison showed that Pikps-1 is 99% identical to Pikp-1, while Pikps-2 is identical to Pikp-2. Pikps-1 had one amino acid substitution (Pro351Ser) in the NBS domain as compared to Pikp-1. The recognition specificity of Pikps against known AVR-Pik alleles is identical to that of Pikp.
Rice (Oryza sativa L.) is a major staple food crop for more than 50% of the world’s population. However, its full production capacity has not been attained due to attacks by pathogens such as viruses, bacteria and fungi. Pathogens secrete effector molecules to enable invasion of the host plant (Hogenhout et al., 2009). To protect themselves, plants have evolved a surveillance system that detects pathogen effectors in the cytosol, leading to effector-triggered immunity (ETI) (Jones and Dangl, 2006). ETI is activated in a gene-for-gene manner upon perception of a specific pathogen effector (AVR) by a cognate Resistance (R) protein in the plant.
Blast disease caused by a filamentous fungus, Pyricularia oryzae (syn. Magnaporthe oryzae), is one of the major diseases, accounting for approximately 30% of global production loss of rice (Nalley et al., 2016). Breeding and deployment of resistant cultivars is the most economical way of controlling the diseases. To date, more than 500 quantitative trait loci (QTLs) for blast resistance have been mapped (Ashkani et al., 2014) and 120 of them have been identified as R-genes (Kalia and Rathour, 2019). Amongst these, 31 R-genes have been cloned and characterized at the molecular level (Xiao et al., 2020). Twenty-eight cloned R-genes, except for pi21 (Fukuoka et al., 2009), Pid2 (Chen et al., 2006) and Ptr (Zhao et al., 2018), encode nucleotide binding site (NBS) leucine-rich repeat (LRR) domain-containing proteins (NLR). These rice NLR genes play a pivotal role in ETI, by recognizing the corresponding AVR gene, and trigger resistance, frequently culminating in the hypersensitive response (HR) leading to cell death (Kourelis and van der Hoorn, 2018). So far, 24 AVR genes of P. oryzae have been mapped and 12 of them have been cloned. In four of the NLR-AVR gene pairs that have been characterized, Pita/AVR-Pita (Jia et al., 2000; Orbach et al., 2000), Pik/AVR-Pik (Ashikawa et al., 2008; Yoshida et al., 2009; Yuan et al., 2011; Zhai et al., 2011; Kanzaki et al., 2012), Pia/AVR-Pia (Yoshida et al., 2009; Okuyama et al., 2011; Ortiz et al., 2017) and PiCO39/AVR1-CO39 (Cesari et al., 2013), the NLR and AVR proteins show direct physical interactions. In the Pii/AVR-Pii pair, however, another host protein, OsExo70, is required for their interaction (Takagi et al., 2013; Fujisaki et al., 2015).
The three classical examples of paired NLR genes of rice, namely the Pik (Pikp-1 and Pikp-2), Pia (RGA4 and RGA5) and Pi5 (Pi5-1 and Pi5-2), also known as Pii (Pii-1 and Pii-2), alleles, are genetically linked in head-to-head orientation (Ashikawa et al., 2008; Lee et al., 2009; Okuyama et al., 2011; Maqbool et al., 2015). One of the paired NLRs (Pikp-1, RGA5 and Pii-2) has a non-canonical domain called the integrated domain (ID), which shares amino acid sequence similarity with domains in other rice proteins and possibly functions as a decoy for recognizing AVR (Cesari et al., 2014a; Kroj et al., 2016; Sarris et al., 2016). These proteins are called “sensor NLRs” because they detect AVRs, whereas the other NLR plays a role in signaling (the “helper NLRs”). The sensor NLRs (Pikp-1 and RGA5) have a heavy-metal-associated (HMA) domain as an ID to which their respective AVRs (AVR-PikD and AVR-Pia) bind directly (Cesari et al., 2014b; Białas et al., 2018).
Recently, extensive molecular studies on rice blast have made it a predominant model pathosystem to understand host–pathogen interactions (Liu et al., 2010). The allelic series of R-genes provide a good platform for deep understanding of the molecular mechanisms involved and the genetic basis of the resistance specificity of the R-genes. The genomic regions harboring allelic series of R-genes are a great resource for comparative study to understand the genomic organization and evolution of R-genes (Zhou et al., 2007). One such region is the Pik locus on chromosome 11 of rice. At the Pik locus, several alleles including Pikm (Ashikawa et al., 2008), Pik (Zhai et al., 2011) and Pikp (Yuan et al., 2011) have been identified. From the pathogen side, four cognate AVR-Pik alleles (AVR-PikA, C, D and E) have been identified and isolated from P. oryzae (Yoshida et al., 2009; Kanzaki et al., 2012). These AVR-Pik alleles are specifically recognized by different Pik alleles. For example, AVR-PikD is recognized by Pik, Pikm and Pikp, while AVR-PikE is recognized by both Pikm and Pik, but AVR-PikA is recognized only by Pikm. AVR-PikC is recognized by none of the Pik alleles reported so far. This gene-for-gene relationship is explained by the NLR-AVR gene co-evolution model (Kanzaki et al., 2012).
In the current study, we found that Oryza sativa subgroup aus cv. Shoni shows resistance against four P. oryza strains. By a QTL mapping and candidate gene cloning strategy, we identified an allele of the rice blast resistance gene Pik and named it Pikps. Rice blast resistance evaluation of cloned genes indicated that like other Pik alleles, Pikps also consists of two NLR genes (Pikps-1 and Pikps-2). Characterization of Pikps indicated that its resistance spectrum and race specificity are similar to those of the Pikp allele of the Pik locus.
The rice cultivar Shoni (WRC31) belongs to O. sativa subgroup aus, and is a member of the world rice collection (WRC) reported by Kojima et al. (2005). Shoni shows resistance to four rice P. oryzae isolates (Supplementary Fig. S1). To identify the resistance gene of Shoni, we used the isolate Naga69-150 (MAFF305471; race code 007.-) as the pathogen. Spray inoculation tests were conducted for the parents (Hitomebore and Shoni) and 125 recombinant inbred lines (RILs) of the F9 generations were developed by a cross between them. Disease symptoms of the RILs were observed eight days post inoculation (dpi) and their disease severity index (DI) scores were categorized into the three classes: no symptoms (DI = 0), with resistance similar to Shoni; 0–20% infected leaf area (DI = 1); and over 20% infected leaf area (DI = 2), with susceptibility similar to Hitomebore (Fig. 1A). Based on the distribution of the average DI of two technical replications of the 125 RILs, we classified RILs with DI below 0.99 (n = 58) and over 1.0 (n = 67) as resistant and susceptible, respectively (Fig. 1B). The observed segregation ratio was tested for goodness of fit to test for a trait controlled by a single locus using chi-square analysis. The expected segregation ratio of a single locus for the RIL population is 1:1 (Liu et al., 2014) and the observed segregation ratio was nearly 1:1 (chi-square, P = 0.65), which indicates that Shoni has a single resistance gene for the P. oryzae isolate Naga69-150.
Resistance assay of 125 RILs against the Naga69-150 isolate of Pyricularia oryzae. (A) The disease severity index (DI) was employed in evaluating phenotypes of RILs after spray inoculation of the fungus. DI = 0, no symptoms; DI = 1, 0 to 20% infected leaf area; DI = 2, over 20% infected leaf area. Representative leaves for each category are shown. Scale bars, 0.5 cm. (B) Frequency distribution of the DI for 125 RILs derived from a cross between Hitomebore and Shoni. Arrows indicate approximate value obtained for the parental (Hitomebore and Shoni) lines. The DI score of each RIL is represented by the average value of two technical replications.
QTL analysis was performed using SNP data from whole-genome sequences of 125 RILs and the DI scores of the inoculation assay. We identified a total of 1,580,242 SNPs between the genomes of the two parents Hitomebore and Shoni. We selected one SNP per 5-kb interval and used 63,551 SNPs for subsequent QTL analysis. QTL analysis was carried out using 125 RILs by the R package GWASpoly (Rosyara et al., 2016) to detect SNPs associated with the observed blast resistance. We identified a single QTL showing statistical significance, i.e., -log10(P) > 3.36, at the end of chromosome 11 (Fig. 2A), which was tentatively named Pi-Shoni. We focused on the region (-log10(P) > 20) corresponding to the position chr11: 27,490,669 to the terminus of chromosome 11 of the Nipponbare reference genome (Kawahara et al., 2013), within which candidate genes were searched. We found a total of 230 protein-coding genes within this region, 11 of which encode NBS-LRR domain-containing proteins (Fig. 2B; Supplementary Table S1). We performed NECAT (https://github.com/xiaochuanle/NECAT/) assembly of the Shoni genome using Nanopore and Illumina sequence reads to compare it with the Nipponbare reference genome (Supplementary Table S2). Genome-wide dot-plot analysis using D-GENIES (Cabanettes and Klopp, 2018) indicated that the Pi-Shoni region of Nipponbare had synteny with the bctg00000014 scaffold (Supplementary Fig. S2). Using exonerate (protein2genome; http://www.ebi.ac.uk/~guy/exonerate), 9 out of 11 genes annotated by the Rice Genome Annotation Project (http://rice.uga.edu) (LOC_Os11g45620, LOC_Os11g45750, LOC_Os11g45790, LOC_Os11g45930, LOC_Os11g45980, LOC_Os11g46200, LOC_Os11g46210, LOC_Os11g47447 and LOC_Os11g47780) were conserved in the bctg00000014 scaffold generated from the Shoni genome assembly (Fig. 2B; Supplementary Table S1). Two of these genes, LOC_Os11g46200 and LOC_Os11g46210, are homologs of Pikm-1 and Pikm-2, both of which are required for the resistance mediated by the rice blast resistance gene Pikm (Ashikawa et al., 2008). We named the homologs of LOC_Os11g46200 and LOC_Os11g46210 in Shoni Pikps-1 and Pikps-2, respectively. We considered these two genes as candidates in Pi-Shoni for conferring resistance against P. oryzae isolate Naga69-150.
Identification of Pi-Shoni (Pikps) conferring resistance on Shoni against Naga69-150. (A) QTL analysis of DI scores obtained from the 125 RILs. The dashed line indicates the significance threshold (-log10(P) > 3.36). (B) Comparative genomic mapping of the 11 NLR genes in Nipponbare within the scaffold bctg00000014 of the Shoni genome assembly. The black arrows indicate NLR genes. Pikps-1 and Pikps-2 correspond to LOC_Os11g46200 and LOC_Os11g46210, respectively.
To verify the functionality of the Pik allele Pikps, we knocked down the Pikps-1 and Pikps-2 genes by the RNA interference (RNAi) method. One of the RILs, RIL#43, carrying Pikps and with suitable transformation efficiency was used for RNAi-mediated gene silencing of Pikps-1 and Pikps-2 (Supplementary Fig. S3). To be confident that Pikps knockdown really caused the change of phenotype, we designed two gene silencing constructs for each gene targeting the coiled-coil (CC) and LRR domains. A total of 76 independent calli were developed after Agrobacterium-mediated transformation of RIL#43 with the four constructs. Out of these, 22 and 16 lines were generated with constructs targeting the CC and LRR domains of Pikps-1, respectively, and 21 and 17 lines were generated for the CC and LRR domains of Pikps-2, respectively. To confirm the function of Pikps-1 and Pikps-2, we carried out punch inoculation of P. oryzae isolate Naga69-150 onto these transformants. Silencing of either of the genes resulted in the compatible reaction, which indicated that both Pikps-1 and Pikps-2 genes are required for Pikps-mediated resistance (Fig. 3A). Reduction of transcript levels of Pikps-1 and Pikps-2 genes was confirmed by quantitative RT-PCR (qRT-PCR) (Supplementary Fig. S4). This result is in support of previous data showing that the Pikm locus on chromosome 11 requires two NLR genes (Pikm-1 and Pikm-2) to manifest disease resistance against rice blast (Ashikawa et al., 2008).
The Pikps-1 and Pikps-2 genes of the cultivar Shoni are responsible for its resistance against Naga69-150. (A) Gene silencing of Pikps-1 and Pikps-2 in the line RIL#43. Numbers below the leaves indicate the callus number. Scale bars, 0.5 cm. (B) Reaction of rice cultivars Hitomebore (Pikp-, Pikm-), RIL#43 (Pikps), K60 (Pikp) and Kanto51 (Pikm) against two AVR-Pik alleles, AVR-PikD and AVR-PikE. The photographs were taken 10 days after inoculation. WT, wild type Sasa2 isolate of P. oryzae. Scale bars, 0.5 cm.
Pikps-1 shared 99% amino acid sequence identity to Pikp-1 with a single amino acid change from serine to proline at position 351 in the NBS domain (Supplementary Fig. S5). Furthermore, Pikps-1 shared 95% amino acid identity with two other Pik alleles, Pik-1 and Pikm-1. On the other hand, Pikps-2 was 100% identical to Pikp-2 and 99% identical to Pik-2 and Pikm-2 (Supplementary Table S3). Pik-1, Pikp-1 and Pikm-1 proteins are known to act as sensor NLRs that bind corresponding AVR-Pik allele effector protein via the integrated HMA domain, whereas Pik-2, Pikp-2 and Pikm-2 are helper NLRs required for activation of the immune response upon effector recognition. Pikps consists of Pikps-1 sensor NLR with an integrated HMA domain and Pikps-2 helper NLR, which is presumably required for initiating resistance signaling. There are no amino acid differences between Pikp-1 and Pikps-1 within the integrated HMA domain (Supplementary Fig. S5).
Characterization of Pikps against known AVR-Pik allelesPrevious studies have shown that Pikp recognizes AVR-PikD, but does not recognize AVR-PikA, C or E (Kanzaki et al., 2012). Since the amino acid sequences of Pikp and Pikps sensor NLR (Pikp-1 and Pikps-1) differ by only one amino acid, it was presumed that Pikps recognizes AVR-PikD and induces resistance in the same manner as Pikp. Gene expression of AVR-PikD was confirmed in rice leaves infected with P. oryzae isolate Naga69-150 (Supplementary Fig. S6). To determine the recognition specificity of Pikps for AVR-Pik alleles, we inoculated RIL#43 with two isolates of P. oryzae (Sasa2 with either the AVR-PikD or the AVR-PikE transgene, as developed by Kanzaki et al., 2012). The wild type Sasa2 isolate does not harbor known AVR-Pik alleles (Yoshida et al., 2009). From the inoculation assay, RIL#43 is incompatible to Sasa2 harboring AVR-PikD, but compatible to Sasa2 as well as Sasa2 harboring AVR-PikE. These results suggest that Pikps recognizes AVR-PikD but not AVR-PikE (Fig. 3B).
In conclusion, by a QTL mapping and cloning strategy, we identified an allele of the rice blast resistance gene Pik, which we named Pikps. Like other Pik alleles, Pikps consisted of two NLR genes (Pikps-1 and Pikps-2). Characterization of Pikps indicated that its race specificity is similar to that of Pikp. Furthermore, Shoni showed resistance against four P. oryzae isolates tested (Supplementary Fig. S1), among which three isolates (Sasa2, 85-141 and Ao92-06-2) do not carry AVR-PikD, as confirmed by the presence/absence of AVR-PikD by PCR amplification and Sanger sequencing of PCR products (Supplementary Fig. S7). These results suggest that Shoni harbors other resistance genes against these three isolates. We addressed the conservation of six well-studied NLR genes (Pia, Pib, Pii, Pit, Pita, Piz-t) in the Shoni genome by BLASTP searches using their amino acid sequences, which showed that Pib and Piz-t are not conserved, while Pia, Pii, Pit and Pita are conserved in Shoni with the amino acid identity ranging from 93.5 to 99.8% (Supplementary Table S4). The race codes of these three strains indicate that they do not have AVR-Pii or AVR-Pia (Supplementary Table S5). Therefore, the resistance of Shoni against the three isolates may be either because of the presence of functional alleles of Pit and/or Pita or other cloned R-genes, as described in Kalia and Rathour (2019), or due to the presence of novel R-genes. Further study is needed to dissect the corresponding R-genes in Shoni for these three isolates.
The japonica-type rice (O. sativa subsp. japonica) cultivar Hitomebore and the aus-type rice (O. sativa subgroup aus) cultivar Shoni (WRC31), provided by the National Agriculture and Food Research Organization World Rice Core Collection (Kojima et al., 2005), were crossed, and 125 RILs of the F9 generation were developed by the single seed descent method. We used the P. oryzae isolates Naga69-150 (race 007, MAFF 305471), 85-141 (race 037.3, MAFF 238767), Ao92-06-2 (race 337.1, MAFF 101530) and Sasa2 (race 037.1, Yoshida et al., 2009) in this study.
Pathogenicity assayA spore suspension (1 × 104 conidia ml−1) of P. oryzae was spray-inoculated onto leaves of 14-day-old plants, which were kept in a dark condition for 24 h at 27 ℃ with 100% relative humidity. The plants were then transferred to a growth chamber with a 16-h light/8-h dark photoperiod. Disease severity was observed visually and scored at 8 days post inoculation with a scale of 0 to 2, as shown in Fig. 1A. For punch inoculation, a conidial suspension (3 × 105 conidia ml−1) was punch-inoculated onto a rice leaf blade one month after seed sowing. The inoculated plants were placed in a dew chamber at 27 ℃ for 24 h in the dark, and transferred to a growth chamber with a 16-h light/8-h dark photoperiod.
Genotyping of RILs by whole-genome resequencingTo obtain the genotypes of the RILs, we performed whole-genome resequencing of the parents and 125 RILs using the Illumina DNA sequencing platform. Genotyping of RILs was carried out according to the method described by Sakai et al. (2021).
Genome assembly of ShoniDNA was extracted from leaf tissue using a NucleoBond HMW DNA kit (Takara Bio, Otsu, Japan). DNA sequencing was performed by Oxford Nanopore Technologies (ONT) using the MinION system with a FLO-MIN106 flow cell (ONT). Base calling of ONT reads was performed on FAST5 files using Guppy (ONT). Subsequently, low-quality reads were filtered out, and de novo assembly was performed using NECAT software (https://github.com/xiaochuanle/NECAT/). To further improve the accuracy of the assembly, Racon software (https://github.com/lbcb-sci/racon) was applied twice, and Medaka (https://github.com/nanoporetech/medaka) was used to correct mis-assembly. One round of consensus correction was performed using BWA (Li and Durbin, 2010) and HyPo (https://github.com/kensung-lab/hypo) on Illumina short reads (trimmed paired-end 150–200-bp reads) for the accession.
RNAi-mediated knock-down of the candidate geneTwo gene knock-down (RNAi) constructs for pANDA-Pi-Shoni candidate genes were generated by PCR amplification of a specific fragment of cDNA of candidate NLR genes from Shoni. The sequences were cloned into the Gateway vector pENTR/D-TOPO (Invitrogen, CA, USA) and transferred into recombination sites of the pANDA vector (Miki and Shimamoto, 2004) using LR Clonase (Invitrogen). One of the 59 RILs that carried Pikps, RIL#43, with a suitable transformation efficiency was selected as the recipient line. The resulting vectors were introduced into Agrobacterium tumefaciens (strain EHA105) and used for Agrobacterium-mediated transformation of rice RIL#43 following the method described by Okuyama et al. (2011). Total RNA was extracted from leaves of transgenic plants using an SV Total RNA Isolation System (Promega, WI, USA) and used for qRT-PCR. cDNA was synthesized from 500 ng total RNA using a PrimeScript RT Reagent Kit (Takara Bio). qRT-PCR was performed using a StepOne Real-time PCR Instrument (Applied Biosystems, CA, USA) with KAPA SYBR FAST PCR Master Mix (Kapa Biosystems, MA, USA). Melting curve analysis (from 60 to 95 ℃) was included at the end of the cycles to ensure the consistency of the amplified products. The comparative Ct (ΔΔCt) method was used to calculate the expression of Pikps-1 and Pikps-2 relative to the rice Actin gene (LOC_Os03g50885) as an internal control. The data presented are the average and standard deviations from three experimental replications. The primers used to generate the RNAi construct and for qRT-PCR are listed in Supplementary Table S6.
This study was supported by JSPS KAKENHI 21K14834 to M. S. and 20H05681 to R. T. We also thank the Genebank at the National Agriculture and Food Research Organization, Japan for providing the seeds of Shoni (WRC31). Pyricularia oryzae isolates Naga69-150 (MAFF 305471), 85-141 (MAFF 238767) and Ao92-06-2 (MAFF 101530) were provided by the Ministry of Agriculture, Forestry and Fisheries, Japan. Computations were partially performed on the NIG supercomputer at ROIS National Institute of Genetics.
