A Natural Allele of a Transcription Factor in Rice Confers ... - Cell Press

Article

A Natural Allele of a Transcription Factor in Rice Confers Broad-Spectrum Blast Resistance Graphical Abstract

Authors Weitao Li, Ziwei Zhu, Mawsheng Chern, ..., Lihuang Zhu, Shigui Li, Xuewei Chen

Correspondence [email protected]

In Brief A natural allele of a C2H2-domain transcription factor gene, bsr-d1, confers broad-spectrum resistance to rice blast.

Highlights d

A single base change (SNP33-G) in the bsr-d1 promoter enhances binding to MYBS1

d

Binding of MYBS1 to the bsr-d1 promoter suppresses bsr-d1 expression

d

BSR-D1 promotes peroxidase expression, suppressing immunity to M. oryzae

d

The SNP33-G allele is present in 10% of 3,000 surveyed rice varieties

Li et al., 2017, Cell 170, 114–126 June 29, 2017 ª 2017 Elsevier Inc. http://dx.doi.org/10.1016/j.cell.2017.06.008

Article A Natural Allele of a Transcription Factor in Rice Confers Broad-Spectrum Blast Resistance Weitao Li,1,6 Ziwei Zhu,1,6 Mawsheng Chern,2,3,6 Junjie Yin,1,6 Chao Yang,1,6 Li Ran,1,6 Mengping Cheng,4 Min He,1 Kang Wang,1 Jing Wang,1 Xiaogang Zhou,1 Xiaobo Zhu,1 Zhixiong Chen,1 Jichun Wang,1 Wen Zhao,1,5 Bingtian Ma,1 Peng Qin,1 Weilan Chen,1 Yuping Wang,1 Jiali Liu,1 Wenming Wang,1 Xianjun Wu,1 Ping Li,1 Jirui Wang,4 Lihuang Zhu,5 Shigui Li,1 and Xuewei Chen1,7,* 1State Key Laboratory of Hybrid Rice, Key Laboratory of Major Crop Diseases and Collaborative Innovation Center for Hybrid Rice in Yangtze River Basin, Rice Research Institute, Sichuan Agricultural University at Wenjiang, Chengdu, Sichuan 611130, China 2Department of Plant Pathology, University of California, Davis, Davis, CA 95616, USA 3Joint Bioenergy Institute, Emeryville, CA 94608, USA 4Triticeae Research Institute, Sichuan Agricultural University, Wenjiang, Chengdu, Sichuan 611130, China 5State Key Laboratory of Plant Genomics and National Center for Plant Gene Research, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing 100101, China 6These authors contributed equally 7Lead Contact *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cell.2017.06.008

SUMMARY

Rice feeds half the world’s population, and rice blast is often a destructive disease that results in significant crop loss. Non-race-specific resistance has been more effective in controlling crop diseases than race-specific resistance because of its broad spectrum and durability. Through a genome-wide association study, we report the identification of a natural allele of a C2H2-type transcription factor in rice that confers non-race-specific resistance to blast. A survey of 3,000 sequenced rice genomes reveals that this allele exists in 10% of rice, suggesting that this favorable trait has been selected through breeding. This allele causes a single nucleotide change in the promoter of the bsr-d1 gene, which results in reduced expression of the gene through the binding of the repressive MYB transcription factor and, consequently, an inhibition of H2O2 degradation and enhanced disease resistance. Our discovery highlights this novel allele as a strategy for breeding durable resistance in rice. INTRODUCTION Despite the lack of a cellular immune system, plants share with animals an innate immune system. Plants contain two major innate immune responses: pathogen-associated molecular pattern (PAMP)-triggered immunity (PTI) (Boller and He, 2009) and effector-triggered immunity (ETI) (Jones and Dangl, 2006). The PTI response includes activation of MAP kinases, induction of reactive oxygen species (ROS), deposition of callose, and induction of pathogenesis-related (PR) genes (Nu¨rnberger et al., 2004). ROS burst constitutes an early response to pathogen 114 Cell 170, 114–126, June 29, 2017 ª 2017 Elsevier Inc.

attack by strengthening cell walls through cross-linking glycoproteins and by activating defense signaling components (Daudi et al., 2012). Pathogens deliver effectors into plant cells to inhibit the host PTI response and/or to create a favoring host cell environment. Plants have developed intracellular sensors encoded by resistance (R) genes containing nucleotide-binding domain-leucinerich repeats (NBS-LRR) that perceive pathogen effectors directly or indirectly leading to ETI (Jones and Dangl, 2006). ETI confers strong resistance, including a hypersensitive response; however, the resistance is limited to a few races of the pathogen and not durable because pathogen effectors evolve quickly (Nu¨rnberger et al., 2004). Because pyramiding R genes to combat diseases is an extremely time-consuming process, deployment of genes conferring broad-spectrum, durable resistance is highly favored by breeders. To date, only six genes conferring non-race-specific resistance have been isolated in plants: Arabidopsis RPW8.2 (Xiao et al., 2001), rice STV11 (Wang et al., 2014), and wheat Lr34 and Yr36 (Fukuoka et al., 2009; Krattinger et al., 2009) are dominant genes; rice pi21 (Fukuoka et al., 2009) and barley mlo (Bu¨schges et al., 1997) act as recessive alleles. Only three have their molecular mechanisms of action elucidated. Arabidopsis RPW8.2, localized to the plant extrahaustorial membrane (EHM), activates the salicylic acid (SA)-dependent defense response, leading to broad-spectrum resistance against powdery mildew (Wang et al., 2009). Rice STV11, encoding a sulfotransferase, catalyzes the conversion of salicylic acid (SA) to sulphonated SA (SSA), leading to durable resistance to rice stripe virus (RSV) (Wang et al., 2014). Wheat Yr36, encoding a kinase-START protein, phosphorylates the thylakoid-associated ascorbate peroxidase (tAPX) in chloroplast, resulting in broad-spectrum resistance to stripe rust (Gou et al., 2015). Although some of these genes have been applied, their deployment to elite cultivars has been limited because of their close linkage to undesirable agricultural traits. For examples, wheat

Figure 1. Identification of Digu-Specific SNPs that Co-segregate with Blast Resistance Using Genome-wide Association Study (A) Chromosomal distribution of Digu-specific SNPs in red. SNPs were identified via comparing Digu with 66 accessions carrying no broad-spectrum resistance. Numbers on rulers represent physical distances in Mb. (B) Co-segregation analysis on SNP33 with the blast resistance using RILs. A total of 12 resistant and 12 susceptible RILs were used for analysis by SNP33. L/D, RILs generated from female parent LTH and male parent Digu; D/L, RILs derived from a reciprocal cross. (C) Structure of LOC_Os03 g32230 and location of SNP33. See also Figures S1, S2, S6, and S7 and Tables S1–S4.

RESULTS

Lr34 lines produce less grain than those without Lr34 (Chen et al., 2016a); the recessive barley mlo mutant causes early senescence-like leaf chlorosis (Piffanelli et al., 2002). New nonrace-specific genes carrying no detrimental effects and not linked to undesirable loci remain to be discovered. Rice blast caused by Magnaporthe oryzae (M. oryzae) is the most devastating disease and greatly reduces yield and grain quality (Dean et al., 2012). Despite recent advances in establishing a good rice-M. oryzae pathosystem, most such reports focus on R-gene-mediated resistance and interaction with M. oryzae. All blast resistance (called Pi) genes encode NBS-LRR proteins, except Pid2 and Pi21 genes. Pid2 encodes a b-lectin receptorlike kinase (Chen et al., 2006). Pi21 encodes a proline-rich protein containing a metal-binding domain and a loss-of-function allele (pi21) confers non-race specific, durable resistance (Fukuoka et al., 2009). These characteristics and the fact that it does not affect yield or grain quality make pi21 a good candidate. Unfortunately, its application is limited because pi21 is closely linked to a gene (LOC_Os04 g32890) that causes inferior grain quality (Fukuoka et al., 2009). Digu is a rice variety carrying durable, high-level resistance to a broad-spectrum of M. oryzae races (Chen et al., 2006; Li et al., 2016), rendering it an excellent resource for discovery of novel genes conferring broad-spectrum resistance. Here, we report our identification of a novel allele, bsr-d1, involved in the broad-spectrum, durable resistance to M. oryzae in Digu. Bsr-d1 encodes a C2H2-type transcription factor, which is directly regulated by a MYB family transcription factor. These two transcription factors regulate expression of H2O2-degradation enzymes to accomplish resistance to M. oryzae, constituting a novel mechanism employed by rice blast resistance.

Genome-wide Association Study Identifies the bsr-d1 Allele as Associated with Digu Blast Resistance To investigate the broad-spectrum resistance to blast in Digu, we carried out genome-wide association study to compare Digu to 66 rice accessions (Table S1) carrying no broad-spectrum resistance randomly selected from the 534 sequenced rice accessions in the China National Crop Gene Bank (CNCGB) (Li et al., 2014). We found 2,576 SNPs that are unique to Digu. Among them, 30 non-synonymous SNPs are located in the exons of 25 genes (Figure 1A; Table S2). Six SNPs are in putative cis-elements in the 1.5-kb promoter regions of six genes (Figure 1A; Table S2). We developed derived cleaved amplified polymorphic sequences (dCAPS) to investigate these SNPs in the 31 identified genes. To facilitate the identification of the responsible SNP, we developed a population of 3,685 recombinant inbred lines (RILs) by crossing Digu with a highly susceptible rice variety, LTH (Figures S1A and S1B). We selected 74 RILs that are morphologically indistinguishable from the LTH parent and that lack both Pid2 and Pid3, two R genes previously isolated from Digu. Among them, 42 are susceptible and 32 resistant to M. oryzae inoculation. We then analyzed the SNPs in ten of the resistant and ten of the susceptible RILs. This initial analysis identified an association between SNP33 and blast resistance (Table S3). We expanded the analysis to all 42 susceptible and 32 resistant lines and found that SNP33 completely cosegregated with blast resistance (Table S3). The blast inoculation results of 12 Digu-type and 12 LTH-type RILs are shown in Figure 1B. These results indicate that the Digu allele (LOC_Os03 g32230) hosting SNP33 (SNP33-G) is tightly associated with the broad-spectrum resistance. SNP33-G is located in the promoter of LOC_Os03 g32230 (Figure 1C). We therefore named it broad-spectrum resistance Digu 1 (bsr-d1). PCR-based sequencing revealed 21 single-base changes and two Indels in

Cell 170, 114–126, June 29, 2017 115

Figure 2. Function of Bsr-d1 on Blast Resistance (A) Blast resistance of Bsr-d1RNAi plants using spraying inoculation. Two inoculated leaves each of Bsr-d1RNAi #1 and #2, and TP309 are shown. Fungal growth, measured as MoPot2 by qRT-PCR, in the inoculated leaves was normalized to OsUbq DNA. (B) Punch inoculation of Bsr-d1RNAi plants. Two leaves each of Bsr-d1RNAi #1 and #2, and TP309 are shown. Leaves of 4-week-old plants were inoculated. Lesion length was measured and fungal growth determined as in (A). (C) Blast susceptibility of Bsr-d1ox plants. Two leaves each of Bsr-d1ox #1 and #2, and TP309 are shown. Lesion length and fungal growth were determined on inoculated leaves at 5 dpi. Leaves of 4-week-old plants were punch-inoculated. (D) Blast resistance of Bsr-d1KO plants. Two leaves each of Bsr-d1KO #1 and #2, and TP309. Lesion length and fungal growth 6 dpi were measured as in (C). In (A)–(D), blast isolate ZB15 was used for inoculations. Error bars represent SEM from three replications. Asterisks represent significant differences (*p < 0.01). See also Figures S2, S3, and S6.

the promoter regions but no differences in the CDSs of Bsr-d1 between Digu and LTH (Figures S1C and S1D).

may have developed a counter strategy to suppress induction of Bsr-d1 expression by M. oryzae.

M. oryzae Induces Bsr-d1 Expression in Susceptible Rice Cultivars, but Not in Resistant Rice To investigate the role of bsr-d1 in immunity to M. oryzae, we assess its expression levels in different rice lines and cultivars before and after blast infection. Interestingly, while the Bsr-d1 RNA level in LTH is only 2-fold higher than in Digu without blast inoculation, it is further induced 4-fold in LTH, but not in Digu, on blast inoculation, rendering the Bsr-d1 RNA level in LTH almost 8-fold compared to that in Digu on blast infection (Figure S2A). We also assessed the Bsr-d1 RNA levels of five Digu-type and five LTH-type RILs. Our results show that the Digu-type lines express low Bsr-d1 RNA levels similar to Digu and are not greatly induced by blast inoculation, whereas all the LTH-type lines display higher Bsr-d1 RNA levels similar to LTH and are greatly induced by blast inoculation despite certain variations (Figure S2A). These results indicate that the Bsr-d1 gene is associated with susceptibility to blast and is induced by blast inoculation in LTH, but not in Digu. To further confirm the inducible nature of Bsr-d1 expression by blast infection and its contribution to blast susceptibility in susceptible rice cultivars, we assessed the Bsr-d1 RNA levels in another three rice cultivars carrying the Digu allele and five cultivars carrying the LTH allele, in addition to Digu and LTH (Figures S2B and S2C). We found that the Bsr-d1 RNA level is induced in all five LTH-allele cultivars (Figure S2C); in contrast, bsr-d1 expression is suppressed in all Digu-allele cultivars (Figure S2B). All three Digu-allele cultivars have been tested resistant and five LTH-allele cultivars susceptible to the blast isolate. These results indicate that M. oryzae induces Bsr-d1 expression, possibly as a strategy to suppress host immunity; rice cultivars, such as Digu,

Silencing, Overexpression, and CRISPR-Mediated Knockout of Bsr-d1 Validate the Role of Bsr-d1 in Digu Resistance In order to further assess the involvement of the Bsr-d1 gene in Digu resistance to blast, we generated transgenic rice plants that either had the Bsr-d1 gene silenced via RNA interference (Bsr-d1RNAi) or overexpressed. We obtained eight putative Bsr-d1RNAi lines in the TP309 rice cultivar and tested their Bsr-d1 RNA levels. Bsr-d1RNAi lines #1 and #2 showed clearly reduced Bsr-d1 RNA levels
35% of wild-type TP309 (Figure S2D). In order to exclude the possibility of cross silencing, we tested the RNA levels of a rice gene (LOC_Os03 g32220) closely related to Bsr-d1 in nucleotide sequence (Figure S2E) in Bsr-d1RNAi #1 and #2 and found no reduction in its RNA levels (Figure S2F). Bsr-d1RNAi lines #1 and #2 were inoculated with blast along with TP309 by spraying with conidia. The two Bsr-d1RNAi lines develop only small, scattered lesions compared to TP309 which shows typical blast lesions (Figure 2A), indicating that silencing of Bsr-d1 leads to enhanced resistance to blast. The inoculation was repeated with a punch method using detached leaves in a controlled environment (see the STAR Methods). Bsr-d1RNAi #1 and #2 developed smaller lesions in the punch inoculation (Figure 2B). The lesion lengths on the Bsr-d1RNAi lines are reduced to 30%–35% compared to that of TP309 (Figure 2B, middle panel). Fungal DNA quantification shows that the Bsr-d1RNAi lines harbor only
30% M. oryzae compared to TP309 (Figure 2B, right panel). This blast resistance phenotype cosegregates with the presence of the transgene when assessed in segregating progeny (Figures S2G and S2H).

116 Cell 170, 114–126, June 29, 2017

We generated two Bsr-d1 overexpression (Bsr-d1ox) lines in the TP309 background and confirmed their elevated Bsr-d1 RNA levels (Figure S2I). We inoculated Bsr-d1ox #1 and #2 lines together with TP309 and found elevated susceptibility to blast (Figure 2C). These two Bsr-d1ox lines develop blast lesions approximately twice as long as TP309 (Figure 2C, middle panel). Fungal DNA quantification also shows that the Bsr-d1ox lines harbor more fungus than TP309 (Figure 2C, right panel). Thus, results of the RNAi and overexpression experiments strongly suggest that the Bsr-d1 gene functions as a negative regulator to blast resistance. In order to further validate the involvement of Bsr-d1 in blast resistance, we used the CRISPR/Cas9 technology to knock out the endogenous Bsr-d1 gene (Bsr-d1KO) in TP309. We selected two 20-nt sequences in the Bsr-d1 gene as target sites for Cas9 cleavage (Figure S3A), generated multiple putative transgenic lines and sequenced the target regions after PCR amplification. We found four mutant lines for target site 1 (named Bsrd1KO #1) and three lines for target site 2 (Bsr-d1KO #2); Bsr-d1KO #1 and #2 each carries a one-base insertion in the target site (Figure S3B), truncating the Bsr-d1 open reading frame. Bsr-d1KO #1 and #2 were punch-inoculated with the ZB15 blast isolate. The two Bsr-d1KO lines develop smaller lesions compared to TP309 (Figure 2D, left panel). The lesion lengths on the Bsr-d1KO lines are reduced to
50% of TP309 (Figure 2D, middle panel). Fungal DNA quantification also shows that the Bsr-d1KO lines harbor only
30% M. oryzae as TP309 (Figure 2D, right panel). Bsr-d1 Knockout Lines, Like Digu, Carry Resistance to a Broad Spectrum of Blast Isolates and Elicit a Hypersensitive Response In order to assess the degree of the involvement of Bsr-d1 in the broad-spectrum resistance to blast in Digu, we tested the resistance of the Bsr-d1KO lines to a collection of nine blast isolates (Zhong10-8-14-eGFP, ZE-1, NC24, B04, NC10, HN41, 9920-2, Guy11, and FJ08-9-1), in addition to ZB15, that Digu confers resistance to. Our results (Figures S3C and S3D) show that the two Bsr-d1KO lines confer enhanced resistance to all nine isolates, reducing lesion lengths to 30%–45% compared to TP309. We next examined growth of a Zhong 10-8-14 isolate tagged by eGFP on sheath cells of TP309 and Bsr-d1KO plants. Appressoria form on TP309 at 12 hpi (hours post inoculation), but not on Bsr-d1KO until 24 hpi (Figure 3A). Invasive hyphae form on TP309 at 24 hpi and extend to the neighboring cells at 36 hpi. In contrast, on Bsr-d1KO plants, invasive hyphae become weakened at 36 hpi and disappear at 48 hpi (Figure 3A). Quantitation of the numbers of M. oryzae at different stages confirms the results (Figure 3B). Therefore, knockout of Bsr-d1 greatly affects M. oryzae growth at the early stage and completely eliminates its growth at later stages. We also examined plant response at the cellular level to investigate the cellular mechanism of the resistance mediated by Bsr-d1 knockout. Staining of the inoculated leaf cells with 3,30 diamino-bezidine (DAB) for H2O2 reveals that the Bsr-d1KO lines produce much higher amounts of H2O2 in the inoculated leaf cells than TP309, resulting in darker staining (Figures 3C and 3D). Consistent with the DAB staining results, the Bsr-d1KO lines display hypersensitive responses 36–48 hpi in the inoculated

sheath cells, but not TP309 (Figure 3A). These results indicate that elevated ROS burst is part of the mechanism of the resistance mediated by Bsr-d1 knockout. BSR-D1 Targets Two Peroxidase Genes Whose Induction by Blast Infection Is Suppressed in Digu Bsr-d1 encodes a putative C2H2-like transcription factor (Figure S4A). To determine whether the BSR-D1 protein is a C2H2type DNA binding protein, we tested if it bound to the EP1S sequence, containing the core C2H2 binding site (Yoshioka et al., 2001) in yeast one-hybrid assay. Indeed, when fused to the GAL4 activation domain (AD), BSR-D1 binds to the EP1S sequence and activates the HIS2 reporter gene (Figure 4A). We carried out subcellular localization of BSR-D1 by fusing it to the GFP protein. The BSR-D1-GFP protein is colocalized to the nucleus with a RFP protein carrying a nuclear localization signal (RFP-NLS) (Figure 4B). Thus, our results suggest that BSR-D1 functions as a transcription factor in the nucleus. In order to assess the target genes of the BSR-D1 protein, we conducted a chromatin immunoprecipitation (ChIP)-seq experiment using GST antibodies against GST-tagged BSR-D1 protein. We identified a list of 14 genes (Table S4). Topping the list of genes pulled down by the BSR-D1 protein are two peroxidase genes (Os05 g04470 and Os10 g39170). The ChIP-seq results were validated by carrying out real-time PCRs to quantitate the presence of these promoters. These promoters are pulled down 2.5- to 4-fold more frequently compared to the control, which had no antibodies added (Figure 4C). Both promoters contain a putative C2H2 binding site (conserved motif) (Figures S4B and S4C). We next validate binding of BSR-D1 to these promoters in the yeast one-hybrid assay, where BSR-D1 was fused to GAL4 AD and each promoter fused to the HIS2 reporter gene. The result shows that BSR-D1 binds to each promoter, leading to activation of the HIS2 reporter (Figure 4D). In conclusion, our results demonstrate that the BSR-D1-GAL4 AD protein binds to the promoter of each of the two peroxidase genes and activates their expression. In order to evaluate whether Bsr-d1 directly regulates expression of the two peroxidase genes, we assessed the expression levels of these peroxidase genes in the Bsr-d1KO lines and found their expression levels are reduced 3- to 8-fold compared to those in TP309 (Figure 4E). In contrast, their expression levels are increased 2- to 3-fold in the Bsr-d1ox lines (Figure 4F). These results indicate that BSR-D1 directly regulates these peroxidase genes. Consequently, we hypothesize that BSR-D1 regulates expression of these two peroxidase genes differentially in Digu and LTH, due to the different Bsr-d1 levels in these two rice varieties, leading to enhanced resistance in Digu, but not in LTH. To test this hypothesis, we assessed the expression levels of these peroxidases in Digu and LTH. Indeed, we found that the expression levels of these two genes are 2- to 8-fold lower in Digu than in LTH (Figure 4G). We further validated the involvement of Os10 g39170, which shows greater induction by Bsr-d1 overexpression in TP309 and by M. oryzae infection in LTH (Figures 4F and 4G). Knockout plants of Os10 g39170 (Os10 g39170-KO) (Figures S4D and S4E) exhibit enhanced resistance whereas overexpression plants of Os10 g39170 (Os10 g39170ox) (Figure S4F) display enhanced susceptibility to M. oryzae compared

Cell 170, 114–126, June 29, 2017 117

Figure 3. Cellular Responses to M. oyrzae Infection (A) Representative laser scanning microscopy images of TP309 and Bsr-d1KO sheath cells infected by eGFP-tagged blast isolate Zhong 10-8-14. Scale bars, 50 mm. (B) Distribution of fungal infection progressing at 12, 24, 36, and 48 hpi. At least 20 single-cell interaction sites were examined per replication. Each bar represents the mean of three replications. Asterisks represent significant differences (p < 0.01) in M. oryzae growth between TP309 and Bsrd1KO plants observed at 12, 24, 36, and 48 hpi, respectively, based on statistical analyses on three replicate experiments (see the STAR Methods). (C) DAB staining at infection sites of Bsr-d1KO and TP309 plants 2 days post inoculation (dpi). Tawny shading indicates accumulation of H2O2. Arrows indicate infection structure of appressoria. Scale bars, 10 mm. (D) Quantitation of H2O2 shown in (C). Relative amount of H2O2 was calculated based on pixels taken with Photoshop, using the formula: H2O2 area per rectangle = pixel of H2O2 area per leaf/ pixel of rectangle. Data are represented as mean ± SEM. Asterisks represent significant differences (*p < 0.01). See also Figure S3.

with TP309 plants (Figures 4H and 4I). These results demonstrate that the level of peroxidase Os10 g39170 is critical for rice resistance to M. oryzae, consistent with the concept that the higher H2O2 level in Digu leads to resistance. The MYBS1 Transcription Factor Binds the Bsr-d1 Promoter Differentially between Digu and LTH Because Bsr-d1 is expressed 6-fold lower in Digu than in LTH on blast infection, and this characteristic is associated with the

118 Cell 170, 114–126, June 29, 2017

resistance phenotype, we looked for transcription factors that may regulate Bsr-d1 expression. Although a total of 21 SNPs between Digu and LTH exist in the Bsr-d1 promoter, only two SNPs are located in predicted transcription factorbinding motifs, both representing binding sites for MYB transcription factors (Figure 5A). Based on these data, we hypothesized that MYB transcription factors might regulate Bsr-d1 expression in Digu differentially from in LTH. We therefore searched for possible MYB transcription factors that are coregulated with Bsr-d1 expression and that may bind to and regulate the Bsr-d1 promoter. We found four MYB genes upregulated and one downregulated after blast inoculation in Digu and LTH (Figure S5A). We then expressed each of these MYB proteins individually fused to the GAL4 AD in a yeast one-hybrid system to assess their potential binding to the Bsr-d1 promoter fused to the HIS3 reporter (Figure S5B). The four upregulated MYB candidates did not significantly affect yeast growth. In contrast, the downregulated MYB candidate (Os01 g34060, abbreviated as MYBS1) significantly reduced yeast growth (Figure S5C), indicating that MYBS1 binds to the Bsr-d1 promoter and inhibits its expression. To confirm the binding and to determine where in the promoter MYBS1 binds, we carried out the electrophoresis mobility shift assay (EMSA). We expressed the MYBS1 protein in E. coli and

Figure 4. Cellular Localization and DNA-Binding Ability of BSR-D1 and the Role of BSR-D1-Targeted Peroxidase Gene (A) Binding to EP1S in yeast one-hybrid. Yeast cells were co-transformed with an effector vector containing EP1S fused to a pHIS2 reporter gene and a prey vector encoding Bsr-d1 fused to GAL4 AD. The SD-Trp-Leu-His medium with 30 mM 3-amino-1,2,4-triazole (3-AT) was used to test for HIS2 expression. (B) Subcellular localization. The BSR-D1-GFP and a nuclear marker pSAT6-mCherry-VirD2NLS were co-expressed in rice protoplast cells. Scale bars, 10 mm. (C) In vitro pull down of target DNA by BSR-D1. GST-BSR-D1 or GST alone were incubated with total rice DNA for 4 hr, pulled down, washed, and subjected to qPCR for peroxidase genes Os05 g04470 and Os10 g39170. The fold enrichment was normalized against the ubiquitin promoter. Each bar represents the mean and SEM of three repeats. *p < 0.01. Similar results were obtained from three independent biological experiments. (D) Bing to two peroxidase promoters in yeast one-hybrid. Each peroxidase promoter was fused to the pHIS2 reporter and BSR-D1 fused to GAL4 AD. Yeast cells transformed with the reporter and effector constructs with or without Bsr-d1. (E) RNA expression levels of the peroxidase genes in Bsr-d1KO plants.

(legend continued on next page)

Cell 170, 114–126, June 29, 2017 119

purified it as a glutathione-S-transferase (GST)-fusion protein. For the probe, we synthesized biotin-labeled oligonucleotides containing each of the two putative MYB binding sites, MYS1 and MYS2, present in the Bsr-d1 promoter region (Figure 5A). The EMSA results show that GST-MYBS1 binds MYS1, but not MYS2, and that it has a higher affinity for the Digu MYS1 than for the LTH MYS1. GST alone does not bind to the MYS1 probe (Figure 5B, left panel). In order to confirm the higher binding affinity for the Digu MYS1, the experiment was repeated with three replicates and the quantitated results (Figure 5B, right panel) validated the above observations. The specificity of the binding was further confirmed by including competitors, where wild-type competitors greatly reduced binding to the probe whereas the mutated competitors had little effect (Figure 5C). Together, these results clearly show that the MYBS1 protein binds to MYS1 in the Bsr-d1 promoter and that it has a much higher affinity for the Digu MYS1 than the LTH MYS1. Knockout of Mybs1 Leads to Elevated Bsr-d1 Expression and Enhanced Susceptibility to M. oryzae Our results indicate that the Mybs1 gene may play a role in regulating Bsr-d1 expression and hence resistance to rice blast. To further assess this, we generated two Mybs1 knockout lines (Mybs1-KO #1 and #2) through the CRISPR technology in the TP309 background: one carried a single-base deletion (two lines) and the other a two-base deletion (five lines) (Figures S5D and S5E), both resulting in a frameshift. These knockout lines were punch-inoculated on detached leaves with blast isolate ZB15 together with TP309 control plants. Mybs1-KO #1 and #2 develop lesions approximately twice as large as TP309 leaves 5 days post inoculation. Measurements of M. oryzae DNA confirm the lesion length results (Figure 6A). To further confirm the results, these lines were inoculated with an eGFP-tagged blast isolate Zhong 10-8-14, and sheath cells were examined under laser scanning microscopy. Invasive hyphae clearly grow more in Mybs1-KO #1 and #2 than in TP309 at 34 and 44 hpi. Quantitative measurements of fungal growth progression show the same results (Figures 6B and 6C). In addition, DAB staining shows that Mybs1-KO #1 and #2 lines generate less H2O2 around appressoria than TP309 plants 5 days post infection (Figures 6D and 6E), consistent with their higher susceptibility. We next assessed the expression levels of Bsr-d1 in the two Mybs1 knockout lines. Their Bsr-d1 RNA levels are elevated 4- to 5-fold compared to TP309 (Figure S5F). Consistently, the two BSR-D1 target (peroxidase) genes are expressed 3- to 6-fold higher in the Mybs1-KO lines (Figure S5G). These results

are consistent with our hypothesis in which MYBS1 functions as a negative regulator of Bsr-d1 expression, leading to elevated defense response. However, when the RNA levels of Mybs1 are assessed over a period of 48 hr after inoculation with or without (mock) a blast isolate using both spraying and punch inoculation methods (Figure S5H), the Mybs1 RNA levels first decrease then return back to normal in both Digu and LTH (Figures S5I and S5J); no clear induction by blast infection was detected and no clear differences between Digu and LTH were observed. Consequently, we conclude that the Mybs1 RNA level is not a major regulatory mechanism resulting in the broad-spectrum resistance in Digu, even though Mybs1 was initially identified due to its reverse correlation with Bsr-d1 expression (Figure S5A). DISCUSSION Despite the popularity of the genome-wide association study (GWAS) approach to mine favorable natural alleles associated with important traits in crops, there are few reports of success. Our identification of the novel bsr-d1 allele by GWAS represents one of the few successful GWAS examples. Digu is a selected natural rice variety that has developed effective, durable broad-spectrum resistance to the notorious rice blast pathogen. We have revealed that the novel bsr-d1 allele contributes to the broad-spectrum, durable resistance in Digu. There are significant impacts on both the conceptual advancement and practical application. M. oryzae Employs Bsr-d1 to Suppress Host Immunity, Facilitating Its Pathogenesis Our results demonstrate that Bsr-d1 is induced on blast infection in five susceptible cultivars and five susceptible recombinant lines but is blocked from induction, or even suppressed, in three resistant cultivars and five resistant recombinant lines. We present a model to summarize our hypothesis (Figure 7). When Bsr-d1 expression is induced in susceptible rice cultivars, host immunity to M. oryzae is suppressed because accumulation of H2O2 is inhibited, like in Bsr-d1ox plants. These results indicate that Bsr-d1 is a target for M. oryzae pathogenesis in order to suppress host immunity through degradation of H2O2 generated in the host PTI response. Conceivably, M. oryzae may use effectors to directly or indirectly induce Bsr-d1 expression. However, we have not identified M. oryzae effectors that act directly on the Bsr-d1 promoter. Rice cultivars, like Digu, have developed the bsr-d1 allele to counter M. oryzae pathogenesis because

(F) RNA expression levels of the peroxidase genes in Bsr-d1ox plants. In (E) and (F), qRT-PCR results were normalized with the Ubq5 reference gene. Error bars represent the SEM of three replicates. Similar results were obtained from three independent biological experiments. (G) RNA expression levels of Os05 g04470 and Os10 g39170 in Digu and LTH. Results are normalized with the Ubq5 reference gene. Samples were collected from Digu and LTH at the seedling stage. Error bars represent the SEM of three replicates. Similar results were obtained from three independent biological experiments. (H) Punch inoculation of Os10 g39170-KO plants. Two leaves each of Os10 g39170-KO #1, #2, and TP309 are shown (left panel). Lesion length and relative fungal DNA (middle and right panels) are measured at 6 dpi. (I) Punch inoculation of Os10 g39170ox plants. Two leaves each of Os10 g39170ox #1, #2, and TP309 are shown (left panel). Lesion length and relative fungal DNA (middle and right panels) are measured at 4 dpi. Blast inoculation and measurements were done as in Figure 2. Same leaves were used to measure fungal growth (MoPot2/OsUbq). In both (H) and (I), blast isolate ZB15 was used for inoculations. Error bars represent the SEM of three replications. Asterisks represent significant differences (*p < 0.01). See also Figure S4.

120 Cell 170, 114–126, June 29, 2017

Figure 5. Direct Binding of MYBS1 to the Bsr-d1 Promoter (A) Two putative MYB binding sites in the Bsr-d1 promoter. MYS1 and MYS2 sequences are each marked by a line above. (B) Electrophoretic mobility shift assay (EMSA). MYS1 and MYS2 probe sequences are as marked in (A). Each probe was labeled with biotin at the 50 end. The relative amounts of the bound form of MYS1 of Digu and LTH were quantitated (right). Error bars represent the SEM of three replicates (*p < 0.01). (C) Binding specificity of MYBS1 to MYS1. Wildtype (50 -TTTAAATTTT AACTGATAAATATA-30 ) or mutant (50 -TTTAAATTTTACCTCATAAATATA-30 ) oligonucleotides was used for competitive binding in 10-, 50-, and 100-fold excess of the wild-type probe. See also Figure S5.

the MYBS1 repressor binds to the bsr-d1 promoter blocking its induction by M. oryzae (Figure 7). Thus, we have uncovered a novel molecular mechanism by which M. oryzae suppresses host immunity and a mechanism the host has evolved to fight against M. oryzae. BSR-D1 and MYBS1 Are Novel Transcription Factors that Regulate Plant Immunity Transcription factors that are involved in plant immunity mainly fall into five families: WRKY, ethylene responsive factor/APETALA2 (ERF/AP2), basic-domain leucine-zipper (bZIP), basic helix-loophelix (bHLH), and NAM/ATAF/CUC (NAC) (Seo et al., 2015). In addition, C2H12, NbSPL6, and MYB30, belonging to the CCCH, SPL, and MYB families, are also involved in plant immunity (Seo et al., 2015). Most of the well-characterized transcription factors involved in plant immunity belong to the WRKY family (Cheng et al., 2015; Liu et al., 2016). OsNAC111 (Yokotani et al., 2014) and OsERF922 (Liu et al., 2012) are two relatively well-studied nonWRKY transcription factors. BSR-D1 does not regulate immunity through OsERF922 or OsNAC111, because Bsr-d1KO did

not affect expression of OsERF922 or OsNAC111 (Figure S6A). BSR-D1, containing a C2H2 domain, represents a novel family of transcription factors that regulate plant immunity. MYBS1, a novel member of MYB transcription factors, suppresses bsr-d1 expression in Digu, leading to higher immunity to M. oryzae. Although MYB transcription factors have been reported to regulate other biological processes through modulation of other transcription factors (Nakano et al., 2015), no reports have shown that a MYB transcription factor plays a critical role in plant immunity by modulating expression of another transcription factor. Our work shows an example of how a MYB transcription factor can be involved in plant immunity. bsr-d1 Likely Confers Durable, Broad-Spectrum Resistance by Regulating Peroxide Accumulation, Like in Digu Bsr-d1KO results in H2O2 accumulation at the infection site and, more importantly, resistance to a broad-spectrum of M. oryzae races, similar to Digu. This characteristic is different from the R protein-mediated resistance, which is normally limited to specific races of the pathogen that carry the corresponding Avr protein. Compared to race-specific resistance mediated by R proteins, non-race-specific resistance is usually more durable. For example, the resistance to M. oryzae conferred by pi21 (a non-R protein) is non-race-specific and durable (Fukuoka et al., 2009). We hypothesize that the resistance conferred by bsr-d1 is durable to M. oryzae, like in Digu. Bsr-d1KO plants, like Digu plants, carry immunity to M. oryzae by regulating H2O2 accumulation (Figures 3 and 7). Reactive oxygen species, including H2O2 and peroxides, serve as a secondary messenger in signal transduction, leading to activation

Cell 170, 114–126, June 29, 2017 121

Figure 6. Role of Mybs1 in Blast Resistance (A) Punch inoculation of the Mybs1KO plants. Two leaves each of Mybs1-KO #1, #2, and TP309 are shown (left panel). Blast inoculation and measurements were done as in Figure 2. Pictures were taken 5 dpi (days post inoculation) and leaves used to measure fungal DNA (MoPot2/OsUbq). Error bars represent the SEM of three replications. Similar results were obtained from three biological experiments. (B) Laser scanning microscopy images of TP309 and Mybs1-KO sheath cells infected by eGFP-tagged blast isolate Zhong 10-8-14. Scale bar, 50 mm. (C) Distribution of fungal infection progressing at 10, 21, 34, and 44 hpi. Procedures were carried out as in Figure 3B. Asterisks represent significant differences (p < 0.01) in M. oryzae growth between TP309 and Mybs1-KO plants observed at 34 and 44 hpi. (D) DAB staining at infection sites of Mybs1-KO and TP309 plants. Leaf samples were collected 2 and 5 dpi. Experiments were done as in Figure 3C. Tawny shading indicated accumulation of H2O2. Arrows indicate infection structure of appressoria. Scale bar, 50 mm. (E) Quantification of H2O2 shown in (D). H2O2 was quantified as in Figure 3. Data are represented as mean ± SEM. Asterisks represent significant differences (*p < 0.01). See also Figure S5 and Table S5.

122 Cell 170, 114–126, June 29, 2017

Figure 7. A Model for bsr-d1-Mediated Disease Resistance In Digu (bsr-d1), MYBS1 binds to the Bsr-d1 promoter with high affinity, suppressing Bsr-d1 expression; low BSR-D1 levels in turn downregulate expression of downstream genes including two peroxidases, resulting in accumulation of H2O2 and enhanced resistance to M. oryzae. In susceptible rice varieties (Bsr-d1), like LTH, Bsr-d1 is highly expressed, activating specific H2O2 degradation activities, leading to susceptibility.

tional induction of OsRboh to achieve enhancement of the immune response. Peroxidases, such as catalases and ascorbate peroxidases, degrade H2O2. There are 73 peroxidase genes in Arabidopsis (Delannoy et al., 2004) and 138 in rice (Passardi et al., 2004). Suppression of tobacco catalase 1 or catalase 2 results in accumulation of H2O2, elevated levels of PR1 protein and salicylic acid, and enhanced resistance to tobacco mosaic virus (Takahashi et al., 1997), demonstrating a direct link between higher H2O2 levels and activation of the immune response. Most importantly, our results demonstrate that peroxidase Os10 g39170 is directly involved in resistance to M. oryzae in rice (Figures 4H and 4I).

of gene expression, enzyme activities, and the immune response (Xu et al., 2016). NADPH oxidases, also known as respiratory burst oxidase homologs (RBOHs), are responsible for most ROS production (Foreman et al., 2003). Three (OsRbohA, OsRbohB, and OsRbohE) of the nine RBOHs in rice have been reported to contribute to resistance to M. oryzae by enhancing production of H2O2 (Wong et al., 2007). Expression of these genes is not affected in Bsrd1KO plants (Figure S6B). Bsr-d1 achieves H2O2 accumulation likely by inhibiting its degradation rather than by increasing its production. Thus, bsr-d1 does not rely on the transcrip-

Application of the bsr-d1-Mediated Resistance to the Blast Disease in Breeding Even though typical Pi genes confer strong resistance to M. oryzae, pyramiding of multiple Pi genes to stack up the resistance spectrum to M. oryzae is a very time-consuming task and may not be an effective way to combat the fastevolving rice blast pathogen. Despite a total of more than 28 Pi genes that have been isolated, all Pi genes encode proteins belonging to the NBS-LRR family with the exception of Pid2 and pi21. Pid2 encodes a b-lectin receptor-like kinase (Chen et al., 2006); pi21 encodes a recessive, uncharacterized proline-rich protein containing a metalbinding domain (Fukuoka et al., 2009). Similar to pi21, bsr-d1 also confers non-race-specific resistance to a broad-spectrum of M. oryzae, constituting a superior disease-resistant characteristic for crop breeding. Interestingly, both pi21 and bsr-d1 are identified as a variant allele from natural rice varieties carrying loss of function of a gene repressing the rice immune response. The bsr-d1 gene, however, carries a further advantage over pi21. The pi21 gene is closely linked to an undesirable gene

Cell 170, 114–126, June 29, 2017 123

(LOC_Os04 g32890) conferring poor flavor and grain quality, especially causing chalkiness in grain (Fukuoka et al., 2009). It will take great effort for breeders to separate pi21 and this undesirable gene. In contrast, bsr-d1, located on chromosome 3, is not linked to any genes known to confer undesirable grain quality or flavor. All known QTLs controlling chalkiness are located on chromosomes 1, 2, and 4–12 (Bian et al., 2013; Chen et al., 2016b); none are located on chromosome 3. Our measurements suggest no differences in important agronomical traits between Bsr-d1KO and wild-type cultivar (Figure S6C). Even better, the Bsr-d1KO grains may carry a slightly lower degree of endosperm chalkiness for unknown reason (Figures S6D–S6G). Therefore, bsr-d1 plants represent a better alternative genetic resource for rice resistant breeding. Enhancement of disease-resistance by modification of defense-related genes often comes with compromise in plant growth. For instance, mutations in Arabidopsis SRFR1 or BAK1 result in dwarfism (Bhattacharjee et al., 2011; Chinchilla et al., 2007), and mutations in SPL28 increase resistance to M. oryzae, but lead to reduction in yield and grain quality (Qiao et al., 2010). In contrast, bsr-d1-mediated resistance to blast carries no observable penalty in plant growth or yield (Figure S6C). The main difference is probably due to the fact that the existence of the bsr-d1 allele has been selected for fitness through environmental challenge and breeding. In fact, our further SNP-Calling analysis (Figure S7) of the 3,000 sequenced rice genomes available in the CNCGB and CAAS database (Li et al., 2014) shows the presence of the bsr-d1 allele in 313 rice varieties distributed in 26 countries, particularly in Southeast Asia (Figure S6H; Table S5). These results indicate that this strategy has been selected during the process of breeding in some areas without knowing the allele. However, because it is present in only 10% of the rice varieties, there is great potential in introducing this natural allele into the other rice varieties. STAR+METHODS Detailed methods are provided in the online version of this paper and include the following: d d d

d

KEY RESOURCES TABLE CONTACT FOR REAGENT AND RESOURCE SHARING EXPERIMENTAL MODEL AND SUBJECT DETAILS B Plant strains B Magnaporthe oryzae isolates METHOD DETAILS B Screening Digu-specific SNPs through GWAS B Protein expression in Escherichia coli and purification B Punch inoculation on rice leaves B RNA isolation and qRT-PCR B Subcellular localization B One-hybrid assays in yeast B Electrophoretic mobility shift assay B Microscopic analysis of pathogen infection B H2O2 accumulation B ChIP in vitro and ChIP-Seq B ChIP-qPCR

124 Cell 170, 114–126, June 29, 2017

B

Analysis of ChIP-Seq Data Plasmid construction and plant transformation QUANTIFICATION AND STATISTICAL ANALYSIS B

d

SUPPLEMENTAL INFORMATION Supplemental Information includes seven figures, six tables, and one data file and can be found with this article online at http://dx.doi.org/10.1016/j.cell. 2017.06.008. AUTHOR CONTRIBUTIONS W.L. and X.C. conceived and designed the experiments. W.L., J.Y., Z.Z., C.Y., L.R., M.H., and K.W. performed experiments on phenotypic and biochemical assays. W.L., C.Y., and L.R. worked on the transgenic lines. M. Cheng and Jirui Wang performed SNP analysis. C.Y., Jing Wang, and W.L. performed the experiments on protein localization. L.R., C.Y., K.W., X. Zhou, X. Zhu, Z.C., W.Z., W.C., and J.L. performed plasmids construction. Jichun Wang, P.Q., and Y.W. created the genetic population. B.M., W.W., X.W., P.L., Jirui Wang, L.Z., S.L., and X.C. collected the data. W.L., M. Chern, and X.C. analyzed data and wrote the manuscript. ACKNOWLEDGMENTS We thank Dr. Yangwen Qian from Hangzhou Biogle Co., Ltd for assistance with gene transformation, and Mr. Zhongxu Chen from Triticeae Research Institute at Sichuan Agricultural University for SNP analysis. We thank the National Supercomputer Centre in Guangzhou houses (SUN YAT-SEN University) for granting CPU-time on the Tianhe-2, Navogene Company (Beijing, China) for analyzing ChIP-seq and Hangzhou Biogle Co., Ltd, for gene knockout constructions and transformations. X.C. was supported by the National Science Foundation of China (NSFC) (31171622 and 31571994), the Program for New Century Excellent Talents in University from the Ministry of Education in China (NECT-13-0920), the National Key Research and Development Program of China (2016YFD0100600), and Transgenic Projects from the Chinese Ministry of Agriculture (TPCMA) (2014ZX0800903B). W.L. was supported by the NSFC (31501627) and the Key Project of Sichuan Education Department (15ZA0020). Ma. C. was supported by the DOE Joint BioEnergy Institute funded by the U.S. Department of Energy (DE-AC0205CH11231). J.Y. was supported by the NSFC (31601290). J.W. was supported by the NSFC (31401351). B.M. was supported by TPCMA (2016ZX08001002). Received: February 2, 2017 Revised: May 1, 2017 Accepted: June 7, 2017 Published: June 29, 2017 REFERENCES Bart, R., Chern, M., Park, C.J., Bartley, L., and Ronald, P.C. (2006). A novel system for gene silencing using siRNAs in rice leaf and stem-derived protoplasts. Plant Methods 2, 13. Bhattacharjee, S., Halane, M.K., Kim, S.H., and Gassmann, W. (2011). Pathogen effectors target Arabidopsis EDS1 and alter its interactions with immune regulators. Science 334, 1405–1408. Bian, J.M., Shi, H., Li, C.J., Zhu, C.L., Yu, Q.Y., Peng, X.S., Fu, J.R., He, X.P., Chen, X.R., Hu, L.F., et al. (2013). QTL mapping and correlation analysis for 1000-grain weight and percentage of grains with chalkiness in rice. J. Genet. 92, 281–287. Boller, T., and He, S.Y. (2009). Innate immunity in plants: an arms race between pattern recognition receptors in plants and effectors in microbial pathogens. Science 324, 742–744. Bu¨schges, R., Hollricher, K., Panstruga, R., Simons, G., Wolter, M., Frijters, A., van Daelen, R., van der Lee, T., Diergaarde, P., Groenendijk, J., et al. (1997).

The barley Mlo gene: a novel control element of plant pathogen resistance. Cell 88, 695–705.

Li, H., and Durbin, R. (2010). Fast and accurate long-read alignment with Burrows-Wheeler transform. Bioinformatics 26, 589–595.

Chen, X., Shang, J., Chen, D., Lei, C., Zou, Y., Zhai, W., Liu, G., Xu, J., Ling, Z., Cao, G., et al. (2006). A B-lectin receptor kinase gene conferring rice blast resistance. Plant J. 46, 794–804.

Li, R., Yu, C., Li, Y., Lam, T.W., Yiu, S.M., Kristiansen, K., and Wang, J. (2009). SOAP2: an improved ultrafast tool for short read alignment. Bioinformatics 25, 1966–1967.

Chen, X., Chern, M., Canlas, P.E., Ruan, D., Jiang, C., and Ronald, P.C. (2010). An ATPase promotes autophosphorylation of the pattern recognition receptor XA21 and inhibits XA21-mediated immunity. Proc. Natl. Acad. Sci. USA 107, 8029–8034.

Li, Z.K., Fu, B.Y., Gao, Y.M., Wang, W.S., Xu, J.L., Zhang, F., Zhao, X.Q., Zheng, T.Q., Zhou, Y.L., Zhang, G., et al.; 3,000 rice genomes project (2014). The 3,000 rice genomes project. Gigascience 3, 7.

Chen, H., Iqbal, M., Yang, R.C., and Spaner, D. (2016a). Effect of Lr34/Yr18 on agronomic and quality traits in a spring wheat mapping population and implications for breeding. Mol. Breed. 36, 53.

Li, W., Liu, Y., Wang, J., He, M., Zhou, X., Yang, C., Yuan, C., Wang, J., Chern, M., Yin, J., et al. (2016). The durably resistant rice cultivar Digu activates defence gene expression before the full maturation of Magnaporthe oryzae appressorium. Mol. Plant Pathol. 17, 354–368.

Chen, L., Gao, W., Chen, S., Wang, L., Zou, J., Liu, Y., Wang, H., Chen, Z., and Guo, T. (2016b). High-resolution QTL mapping for grain appearance traits and co-localization of chalkiness-associated differentially expressed candidate genes in rice. Rice (N. Y.) 9, 48.

Liu, D., Chen, X., Liu, J., Ye, J., and Guo, Z. (2012). The rice ERF transcription factor OsERF922 negatively regulates resistance to Magnaporthe oryzae and salt tolerance. J. Exp. Bot. 63, 3899–3911.

Cheng, H., Liu, H., Deng, Y., Xiao, J., Li, X., and Wang, S. (2015). The WRKY45-2 WRKY13 WRKY42 transcriptional regulatory cascade is required for rice resistance to fungal pathogen. Plant Physiol. 167, 1087–1099. Chinchilla, D., Zipfel, C., Robatzek, S., Kemmerling, B., Nu¨rnberger, T., Jones, J.D., Felix, G., and Boller, T. (2007). A flagellin-induced complex of the receptor FLS2 and BAK1 initiates plant defence. Nature 448, 497–500. Cong, L., Ran, F.A., Cox, D., Lin, S., Barretto, R., Habib, N., Hsu, P.D., Wu, X., Jiang, W., Marraffini, L.A., and Zhang, F. (2013). Multiplex genome engineering using CRISPR/Cas systems. Science 339, 819–823. Daudi, A., Cheng, Z., O’Brien, J.A., Mammarella, N., Khan, S., Ausubel, F.M., and Bolwell, G.P. (2012). The apoplastic oxidative burst peroxidase in Arabidopsis is a major component of pattern-triggered immunity. Plant Cell 24, 275–287. Dean, R., Van Kan, J.A.L., Pretorius, Z.A., Hammond-Kosack, K.E., Di Pietro, A., Spanu, P.D., Rudd, J.J., Dickman, M., Kahmann, R., Ellis, J., et al. (2012). The top 10 fungal pathogens in molecular plant pathology. Mol. Plant Pathol. 13, 414–430. Delannoy, E., Marmey, P., Penel, C., and Nicole, M. (2004). The plant peroxidases of class III. Acta Bot. Gallica 151, 353–380. Feng, Z., Zhang, B., Ding, W., Liu, X., Yang, D.L., Wei, P., Cao, F., Zhu, S., Zhang, F., Mao, Y., and Zhu, J.K. (2013). Efficient genome editing in plants using a CRISPR/Cas system. Cell Res. 23, 1229–1232. Foreman, J., Demidchik, V., Bothwell, J.H., Mylona, P., Miedema, H., Torres, M.A., Linstead, P., Costa, S., Brownlee, C., Jones, J.D., et al. (2003). Reactive oxygen species produced by NADPH oxidase regulate plant cell growth. Nature 422, 442–446. Fukuoka, S., Saka, N., Koga, H., Ono, K., Shimizu, T., Ebana, K., Hayashi, N., Takahashi, A., Hirochika, H., Okuno, K., and Yano, M. (2009). Loss of function of a proline-containing protein confers durable disease resistance in rice. Science 325, 998–1001. Gou, J.Y., Li, K., Wu, K., Wang, X., Lin, H., Cantu, D., Uauy, C., Dobon-Alonso, A., Midorikawa, T., Inoue, K., et al. (2015). Wheat stripe rust resistance protein WKS1 reduces the ability of the thylakoid-associated ascorbate peroxidase to detoxify reactive oxygen species. Plant Cell 27, 1755–1770. Hsu, P.D., Scott, D.A., Weinstein, J.A., Ran, F.A., Konermann, S., Agarwala, V., Li, Y., Fine, E.J., Wu, X., Shalem, O., et al. (2013). DNA targeting specificity of RNA-guided Cas9 nucleases. Nat. Biotechnol. 31, 827–832. Jones, J.D.G., and Dangl, J.L. (2006). The plant immune system. Nature 444, 323–329. Krattinger, S.G., Lagudah, E.S., Spielmeyer, W., Singh, R.P., Huerta-Espino, J., McFadden, H., Bossolini, E., Selter, L.L., and Keller, B. (2009). A putative ABC transporter confers durable resistance to multiple fungal pathogens in wheat. Science 323, 1360–1363. Landt, S.G., Marinov, G.K., Kundaje, A., Kheradpour, P., Pauli, F., Batzoglou, S., Bernstein, B.E., Bickel, P., Brown, J.B., Cayting, P., et al. (2012). ChIP-seq guidelines and practices of the ENCODE and modENCODE consortia. Genome Res. 22, 1813–1831.

Liu, J., Chen, X., Liang, X., Zhou, X., Yang, F., Liu, J., He, S.Y., and Guo, Z. (2016). Alternative splicing of rice WRKY62 and WRKY76 transcription factor genes in pathogen defense. Plant Physiol. 171, 1427–1442. Livak, K.J., and Schmittgen, T.D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402–408. Miki, D., and Shimamoto, K. (2004). Simple RNAi vectors for stable and transient suppression of gene function in rice. Plant Cell Physiol. 45, 490–495. Nakano, Y., Yamaguchi, M., Endo, H., Rejab, N.A., and Ohtani, M. (2015). NAC-MYB-based transcriptional regulation of secondary cell wall biosynthesis in land plants. Front. Plant Sci. 6, 288. Nishimura, A., Aichi, I., and Matsuoka, M. (2006). A protocol for Agrobacterium-mediated transformation in rice. Nat. Protoc. 1, 2796–2802. Nu¨rnberger, T., Brunner, F., Kemmerling, B., and Piater, L. (2004). Innate immunity in plants and animals: striking similarities and obvious differences. Immunol. Rev. 198, 249–266. Ouyang, S., Zhu, W., Hamilton, J., Lin, H., Campbell, M., Childs, K., ThibaudNissen, F., Malek, R.L., Lee, Y., Zheng, L., et al. (2007). The TIGR Rice Genome Annotation Resource: improvements and new features. Nucleic Acids Res. 35, D883–D887. Park, C.H., Chen, S., Shirsekar, G., Zhou, B., Khang, C.H., Songkumarn, P., Afzal, A.J., Ning, Y., Wang, R., Bellizzi, M., et al. (2012). The Magnaporthe oryzae effector AvrPiz-t targets the RING E3 ubiquitin ligase APIP6 to suppress pathogen-associated molecular pattern-triggered immunity in rice. Plant Cell 24, 4748–4762. Passardi, F., Longet, D., Penel, C., and Dunand, C. (2004). The class III peroxidase multigenic family in rice and its evolution in land plants. Phytochemistry 65, 1879–1893. Piffanelli, P., Zhou, F., Casais, C., Orme, J., Jarosch, B., Schaffrath, U., Collins, N.C., Panstruga, R., and Schulze-Lefert, P. (2002). The barley MLO modulator of defense and cell death is responsive to biotic and abiotic stress stimuli. Plant Physiol. 129, 1076–1085. Qiao, Y., Jiang, W., Lee, J., Park, B., Choi, M.S., Piao, R., Woo, M.O., Roh, J.H., Han, L., Paek, N.C., et al. (2010). SPL28 encodes a clathrin-associated adaptor protein complex 1, medium subunit micro 1 (AP1M1) and is responsible for spotted leaf and early senescence in rice (Oryza sativa). New Phytol. 185, 258–274. Seo, E., Choi, D., and Choi. (2015). Functional studies of transcription factors involved in plant defenses in the genomics era. Brief. Funct. Genomics 14, 260–267. Takahashi, H., Chen, Z., Du, H., Liu, Y., and Klessig, D.F. (1997). Development of necrosis and activation of disease resistance in transgenic tobacco plants with severely reduced catalase levels. Plant J. 11, 993–1005. Wang, W., Wen, Y., Berkey, R., and Xiao, S. (2009). Specific targeting of the Arabidopsis resistance protein RPW8.2 to the interfacial membrane encasing the fungal Haustorium renders broad-spectrum resistance to powdery mildew. Plant Cell 21, 2898–2913.

Cell 170, 114–126, June 29, 2017 125

Wang, Q., Liu, Y., He, J., Zheng, X., Hu, J., Liu, Y., Dai, H., Zhang, Y., Wang, B., Wu, W., et al. (2014). STV11 encodes a sulphotransferase and confers durable resistance to rice stripe virus. Nat. Commun. 5, 4768.

Yin, J., Zhu, X., Yuan, C., Wang, J., Li, W., Wang, Y., He, M., Cheng, Q., Ye, B., Chen, W., et al. (2015). Characterization and fine mapping of a novel vegetative senescence lethal mutant locus in rice. J. Genet. Genomics 42, 511–514.

Wong, H.L., Pinontoan, R., Hayashi, K., Tabata, R., Yaeno, T., Hasegawa, K., Kojima, C., Yoshioka, H., Iba, K., Kawasaki, T., and Shimamoto, K. (2007). Regulation of rice NADPH oxidase by binding of Rac GTPase to its N-terminal extension. Plant Cell 19, 4022–4034.

Yokotani, N., Tsuchida-Mayama, T., Ichikawa, H., Mitsuda, N., Ohme-Takagi, M., Kaku, H., Minami, E., and Nishizawa, Y. (2014). OsNAC111, a blast disease-responsive transcription factor in rice, positively regulates the expression of defense-related genes. Mol. Plant Microbe Interact. 27, 1027–1034.

Xiao, S., Ellwood, S., Calis, O., Patrick, E., Li, T., Coleman, M., and Turner, J.G. (2001). Broad-spectrum mildew resistance in Arabidopsis thaliana mediated by RPW8. Science 291, 118–120. Xu, J., Meng, J., Meng, X., Zhao, Y., Liu, J., Sun, T., Liu, Y., Wang, Q., and Zhang, S. (2016). Pathogen-responsive MPK3 and MPK6 reprogram the biosynthesis of indole glucosinolates and their derivatives in Arabidopsis immunity. Plant Cell 28, 1144–1162.

126 Cell 170, 114–126, June 29, 2017

Yoshioka, K., Fukushima, S., Yamazaki, T., Yoshida, M., and Takatsuji, H. (2001). The plant zinc finger protein ZPT2-2 has a unique mode of DNA interaction. J. Biol. Chem. 276, 35802–35807. Zhang, Y., Liu, T., Meyer, C.A., Eeckhoute, J., Johnson, D.S., Bernstein, B.E., Nusbaum, C., Myers, R.M., Brown, M., Li, W., and Liu, X.S. (2008). Modelbased analysis of ChIP-seq (MACS). Genome Biol. 9, R137.

STAR+METHODS KEY RESOURCES TABLE

REAGENT or RESOURCE

SOURCE

IDENTIFIER

Magnaporthe oryzae: ZB15

This paper

N/A

Magnaporthe oryzae: ZE-1

This paper

N/A

Magnaporthe oryzae: NC24

This paper

N/A

Magnaporthe oryzae: B04

This paper

N/A

Magnaporthe oryzae: NC10

This paper

N/A

Magnaporthe oryzae: HN41

This paper

N/A

Magnaporthe oryzae: 9920-2

This paper

N/A

Magnaporthe oryzae: Guy11

This paper

N/A

Magnaporthe oryzae: FJ08-9-1

This paper

N/A

Magnaporthe oryzae: Zhong10-8-14-eGFP

This paper

N/A

Rice varieties: Digu

This paper

N/A

Rice varieties: Lijiangxintuanheigu

This paper

N/A

Rice varieties: TP309

This paper

N/A

GST-Bsr-d1

This paper

N/A

GST-Mybs1

This paper

N/A

Bsr-d1-GFP

This paper

N/A

pSAT6:mCherry(RFP):VirD2NLS

This paper

N/A

p35S:GFP

This paper

N/A

Bacterial and Virus Strains

Biological Samples

Chemicals, Peptides, and Recombinant Proteins

Critical Commercial Assays Light Shift Chemiluminescent EMSA Kit

ThermoFisher SCIENTIFIC

Cat#20148

Gateway LR Clonase Enzyme mix

Thermo Fisher SCIENTIFIC

Cat#11791019

pEASY-Blunt Cloning Kit

TRAN

Cat#L20213

Phusion Hot Start II DNA Ploymerase

ThermoFisher SCIENTIFIC

Cat#00241667

PrimeScriptTM RT reagent Kit with gDNA Eraser

TaKaRa

Cat#RR047A

Deposited Data Raw data of Re-sequencing

This paper

NCBI: SRP102176

Raw data of ChIP-seq

This paper

NCBI: SRP102677

Escherichia coli BL21 (DE3)

Takara

D90120-9125 9057

Experimental Models: Organisms/Strains Escherichia coli DH5a

Clontech

Saccharomyces cerevisiae: Y2HGold and Y187

Clontech

Cat#630489 and Cat#630457

Agrobacterium tumefaciens EHA105

This paper

N/A

The biotin end-labeled probes bound by MYBS1: Biotin-TTTAAATTTTAACTGATAAATATA

This paper

N/A

The biotin end-labeled probes bound by MYBS1: Biotin-TTTAAATTTTAGCTGATAAATATA

This paper

N/A

The biotin end-labeled probes bound by MYBS1: Biotin-CCTTATATGCAACTGGCCGCCTAC

This paper

N/A

Oligonucleotides

(Continued on next page)

Cell 170, 114–126.e1–e5, June 29, 2017 e1

Continued REAGENT or RESOURCE

SOURCE

IDENTIFIER

The biotin end-labeled probes bound by MYBS1: Biotin-CCTTATATGCAAGTGGCCGCCTAC

This paper

N/A

Primers for this study, see Table S6

This paper

N/A

Plasmid: GST-Mybs1

This paper

N/A

Plasmid: GST-Bsr-d1

This paper

N/A

Plasmid: Bsr-d1-GFP

This paper

N/A

Plasmid: pHIS2- EP1S

This paper

N/A

Plasmid: pHIS2-the promoter of Bsr-d1

This paper

N/A

Recombinant DNA

Plasmid: pHIS2-the promoter of Os05 g04470

This paper

N/A

Plasmid: pHIS2-the promoter of Os10 g39170

This paper

N/A

Plasmid: pGADT7-Mybs1

This paper

N/A

Plasmid: pGADT7-Bsr-d1

This paper

N/A

Plasmid: pGADT7-Os02 g41510

This paper

N/A

Plasmid: pGADT7-Os04 g43680

This paper

N/A

Plasmid: pGADT7-Os06 g45890

This paper

N/A

Plasmid: pGADT7-Os10 g33810

This paper

N/A

Plasmid: pANDA-Bsr-d1RNAi

This paper

N/A

Plasmid: Ubi-C1300-Bsr-d1

This paper

N/A

Plasmid: Ubi-C1300-Os10 g39170

This paper

N/A

Plasmid: pBGK032-Mybs1

This paper

N/A

Plasmid: pBGK032-Bsr-d1

This paper

N/A

Plasmid: pBGK032-Os10 g39170

This paper

N/A

Software and Algorithms bwa-0.7.12

Li and Durbin, 2010

http://bio-bwa.sourceforge.net/

picard-tools-2.2.1

GitHub

https://broadinstitute.github.io/picard/

GATK 3.6

GitHub

https://software.broadinstitute.org/gatk/

TESS V1.0

eScience Lab team

http://www.esciencelab.org.uk/announcements/ tess/product/2016/08/09/tess_v1.0_release/

phantompeakqualtools

Landt et al., 2012

https://www.encodeproject.org/

CONTACT FOR REAGENT AND RESOURCE SHARING Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Xuewei Chen ([email protected]). EXPERIMENTAL MODEL AND SUBJECT DETAILS Plant strains Recombinant inbred lines (RILs) were developed by crossing japonica variety Lijiangxintuanheigu (LTH) and indica variety Digu. The F2 generation from LTH 3 Digu or Digu 3 LTH was subjected to more than six rounds of self-pollination to generate the RILs. Japonica variety TP309 was used for transformation to create the transgenic lines. The following transgenic lines were used: pANDA-Bsr-d1RNAi, Ubi-C1300-Bsr-d1, pBGK032-Bsr-d1, pBGK032-Os10 g39170, Ubi-C1300-Os10 g39170 and pBGK032Mybs1. The plants were grown in the fields at Sichuan Agricultural University in Wenjiang, Chengdu, or Lingshui, Hainan, China. Magnaporthe oryzae isolates M. oryzae isolates, ZB15, ZE-1, NC24, B04, NC10, HN41, 9920-2, Guy11, FJ08-9-1, and a GFP-tagged M. oryzae isolate Zhong10-814-eGFP, were used for inoculation. All isolates were grown in a growth chamber at 25 C in a 12 hr light / 12 hr dark photoperiod before used for inoculation.

e2 Cell 170, 114–126.e1–e5, June 29, 2017

METHOD DETAILS Screening Digu-specific SNPs through GWAS SNP calling pipelines is shown in Figure S7. The details are as follows: (1) Obtain raw data of rice Digu genome (NCBI: SRP102176) by deep-sequencing via Navogene Company (Beijing, China) and download raw data of the other 66 rice accessions from GigaScience database (http://gigadb.org/dataset/200001). (2) Combine several .fq files of each accession (Script 1). (3) Convert FASTQ to uBAM and add read groups information using FastqToSam (Script 2). (4) Mark Adapters (Script 3). (5) Map reads of each accession to genome reference (Nipponbare) using bwa-0.7.12 (http://bio-bwa.sourceforge.net/) (Script 4). (6) Splice sequences that were identified to reference (Script 5). The picard-tools-2.2.1 (https://broadinstitute.github.io/picard/) was used to convert a FASTQ file to an unaligned BAM file, read BAM file and rewrites it with new adaptor-trimming tags, to merge alignment data from a BAM with data in an unmapped BAM file, and to locate and tag duplicate reads in a BAM file. (7) Mark Duplicates and remove them (Script 6). (8) Call SNPs using GATK 3.6 (https://software.broadinstitute.org/gatk/) (used the Chr1 as an example. Because of large data, every chromosome was analyzed respectively) (Script 7). (9) Merg callSNP results of 12 chromosomes and screened the Digu-specific SNPs. (10) Manage .gtf files. The information of exon, CDS, start-codon and stop-codon were included in .gtf files. The longest transcripts were kept if several candidate transcripts were existed for one gene. Meanwhile, deleted the 50 and 30 sequences before start-codon and after stop-codon of CDS. (11) Screen for Digu-specific SNPs based on the results from the last step. Characterize Digu-specific SNPs in CDS and its promoter regions (up to 1.5 kb). (12) Screen for the SNPs that cause amino acid substitutions. (13) Screen for Digu-specific SNPs in 1.5 kb promoter regions and in the cis-elements bound by transcription factors using TESS V1.0 (http://www. esciencelab.org.uk/announcements/tess/product/2016/08/09/tess_v1.0_release/). The primers of Derived Cleaved Amplified Polymorphic Sequences (dCAPS) were developed to investigate these SNPs (Table S6). Protein expression in Escherichia coli and purification The full-length cDNA sequence of Mybs1 was cloned into the pGEX-6p-1 vector between the EcoRI and XhoI sites to generate Glutathione-S-transferase (GST):Mybs1. The fusion constructs were then introduced into BL21 (DE3) pLysS E. coli. Bacteria contain the plasmids were grown in Luria-Bertani (LB) medium containing 100 mg/mL and Ampicillin at 37 C to OD600 = 0.6. Expression of the fusion protein GST-MYBS1 was induced by addition of 0.5 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) and incubation at 28 C for 14 hr. The GST-MYBS1 was purified using glutathione-agarose beads (BD biosciences) by following the manufacturer’s instructions. Protein expression of GST-BSR-D1 and purification was also carried out using the same method described above. Punch inoculation on rice leaves M. oryzae isolates were gown on complete agar medium for two weeks before producing spores. Spores were collected via flooding of the fungal agar cultures with sterile water, and the spore concentration in the suspension was adjusted to 5 3 105 conidia/mL before punch inoculation. Punch inoculation of detached rice leaves is performed as follows: Dip 4 mL spore suspension for each drop using transferpettor at two spots on each leaf, keep them in a culture dish that contains 0.1% 6-Benzylaminopurine (6-BA) sterile water to keep moist, and measure lesion length using ruler five days post inoculation. Relative fungal DNA amount was calculated using the threshold cycle value (CT) of M. oryzae Pot2 DNA against the CT of rice genomic ubiquitin DNA (Park et al., 2012). All inoculation experiments were repeated three times independently. RNA isolation and qRT-PCR Total RNA was extracted using TRIzol reagent (Invitrogen Life Technologies, Shanghai, China) according to the manufacturer’s protocols. cDNA was synthesized using an RNA reverse transcription kit (Invitrogen Life Technologies, Shanghai, China). The qRT-PCR was conducted using a Bio-Rad CFX96 Real-Time System coupled to a C1000 Thermal Cycler (Bio-Rad, Hercules, CA, USA). The reference gene Ubiquitin 5 (Ubq5) was used for the normalization of all qRT-PCR data (Li et al., 2016). Sequences of the primers are listed in Table S6. The 2-66CT method was used to calculate the relative expression levels with three technique repeats (Livak and Schmittgen, 2001). Subcellular localization For subcellular localization in rice protoplasts, cDNA fragment corresponding to the entire coding sequence of Bsr-d1 was cloned into between BamHI and HindIII sites of the PBI211 vector to generate PBI211-Bsr-d1 (p35S:Bsr-d1:GFP). The fusion construct was transformed or co-transformed into protoplasts prepared from TP309 seedlings following the method as described previously (Bart et al., 2006). For controls, protoplasts were transformed with pSAT6:mCherry (RFP):VirD2NLS and PBI211 (p35S:GFP), respectively. Fluorescence was examined under a confocal microscopy (NiKon A1 i90, LSCM, Japan) at 16 hr after transformation. One-hybrid assays in yeast The full-length cDNA sequences of transcription factor genes (Mybs1, Bsr-d1, Os02 g41510, Os04 g43680, Os06 g45890, and Os10 g33810) were amplified and fused in frame with the GAL4 activation domain of pGADT7-Rec2 (Clontech), respectively, forming pGADT7-TFs. Then, the fusion construct was co-transformed with the reporter vector (pHIS2-cis/promoter of EP1S/Bsr-d1/Os05 g04470/Os10 g39170) into Y187 yeast cells (Clontech). Sequences of the primers are listed in Table S6. The empty vector

Cell 170, 114–126.e1–e5, June 29, 2017 e3

pGADT7-Rec2 and the pHIS2-cis/promoter were co-transformed as the control for mating experiments. DNA-protein interactions were determined by the growth of the transformants on the nutrient-deficient medium with 30 mM 3-amino-1,2,4-triazole (3-AT), following the manufacturer’s manual (Clontech). Electrophoretic mobility shift assay Light Shift Chemiluminescent EMSA Kit (No. 20148, ThermoFisher SCIENTIFIC) was used in this experiment. The GST-MYBS1 protein was purified using glutathione-agarose beads (BD biosciences) after incubation. Biotin was labeled at 50 end of cis-element. The biotin-labeled DNA were synthesized by Sangon Biotech (Shanghai, China). The components of binding reaction include: 1X Binding Buffer, 2.5% Glycerol, 5mM MgCl2, 50 ng/mL Poly (dI$dC), 0.05% NP-40, 1 Unit Protein, 20 fmol biotin-labeled DNA. The detailed procedure of EMSA follows the manufacturer’s instructions. Photos were taken using Charge-coupled device (CCD) camera. Microscopic analysis of pathogen infection The Zhong10-8-14-eGFP isolate was used to inoculate detached rice sheaths form 4-week-old rice plants. The fungus stably expressing GFP was grown on complete agar medium for two weeks before producing spores. Spores were collected via flooding of the fungal agar cultures with sterile water, and the spore concentration was adjusted to approximately 5 3 105 conidia/ mL. The detached rice sheath assay was performed as described previously (Li et al., 2016). All images of conidia germination, appressorium development and invasive hyphae growth were recorded using a fluorescent microscope (Zeiss imager A2). For growth stage quantitation, the numbers of M. oryzae cells at different stages, including no appressoria formed, appressoria formed, those infected one, two, three, and more plant cells, and dead M. oryzae cells, were counted at each of four time points. Each number was then converted into percentage of total cells. The percentages of M. oryzae cells at each stage on transgenic plants at each time point from three independent replicate experiments were compared to those at the same stage on TP309 plants to analyze for statistical significance using One-way ANOVA. Significant differences are presented only when all growth stages at that time point show significant differences. H2O2 accumulation For cellular response of rice to M. oryzae infection, the M. oryzae isolate ZB15 was inoculated on rice leaves. H2O2 accumulation was monitored via staining with DAB as described by a previous report (Yin et al., 2015). Leaf sections were placed in 1 mg mL-1 DAB (Sigma) and incubated at 22 C for 10 hr at illumination. The DAB-stained leaves were observed under a microscope (Zeiss imager A2). ChIP in vitro and ChIP-Seq Total DNA of Nipponbare and purified GST-BSR-D1 were used for ChIP assays. Three-week-old seedlings were used for total DNA extraction. The total DNA was sheared into 100-500bp fragments using ultrasonic crusher. The GST fusion protein was affinity-purified on glutathione-agarose beads (BD biosciences). GST-BSR-D1 and DNA fragments were co-incubated for 2h. The incubation buffer includes: 50 mM Tris, 1mM EDTA, 100 mM KCl, adjust pH to 7.0 by HCl, 5% Glycerol, 0.1% Triton X-100; add freshly-made 100 mM DTT to reaction solution to make final concentration of DTT at 1mM. After co-incubation, glutathione-agarose beads were washed three times using incubation buffer. Then 4 mL 5M NaCl was added into the sample for each 100 mL volume and was incubated for 4 hr to break down cross-linked GST-BSR-D1 and DNA fragments. Subsequently, DNA fragments were extracted using the phenol-chloroform methods for constructing Illumina sequencing library. Illumina sequencing library was performed with and aboveprepared DNA samples. The ends of DNA fragments were repaired and ligated to an adaptor (AATGATACGGCGACCACCGAGATC TACACTCTTTCCCTACACGACGCTCTTCCGATCT). Then DNA fragments between100 and 500 bp were recovered from the gel and amplified by PCR for 20 cycles. The amplified DNA products were collected, ligated to the PEASY-Blunt vector for a quality test, and then sequenced (NCBI: SRP102677) with Hiseq2500 (Novogene, Beijing). ChIP-qPCR The prepared DNA in ChIP was applied for qPCR using respective primer pairs (Table S6) in an EVER Green PCR Master Mix (BioRad) with a Bio-Rad CFX96 real-time PCR detection system. PCR reactions were performed in triplicate for each sample, and the expression levels were normalized to the input sample for enrichment detection. The fold enrichment was calculated against the Ubq5 promoter. No addition of antibodies (NoAbs) served as a negative control. Analysis of ChIP-Seq Data Sequencing reads for ChIP and input DNA were mapped to Release 7 of Michigan State University Rice Genome (Ouyang et al., 2007) using SOAP2 with parameters -r 0 -n 0 -v 0 -l 75 (Li et al., 2009). Cross-correlation metrics were calculated using phantompeakqualtools (https://encodeproject.org/ENCODE/encodeTools.html#metrics) (Landt et al., 2012). Only uniquely mapped reads were used for peak identification, and BSR-D1 binding peaks were obtained by model-based analysis of ChIP-seq with default parameters (Zhang et al., 2008). The peak summits were used to define the location types in the genome by the following criteria. If a peak summit was located in (1) a gene’s first intron, (2) a gene’s promoter region (upstream 2000 bp from the transcription start sites (TSS), (3) a gene’s exon, or (4) a gene’s intron, it was labeled according to the first criterion it matched. Because of the detailed peak distribution

e4 Cell 170, 114–126.e1–e5, June 29, 2017

on the genome and promoter region, a strict cutoff was used to identify BSR-D1 target genes by labeling the genes with peak summits located in the upstream 1000 bp from the TSS. Plasmid construction and plant transformation For RNAi construct, a unique cDNA fragment of Bsr-d1 from 630 to 957 bp after ATG codon was cloned into the pANDA vector (Miki and Shimamoto, 2004) to create the RNA interference (RNAi) construct, pANDA-Bsr-d1RNAi, using LR recombination enzyme (Invitrogen, Carlsbad, CA, USA). The full-length cDNA of Bsr-d1 was cloned into pENTR/D to make pENTR-Bsr-d1, which was then recombined with the Ubi-C1300GTW vector to generate the overexpression construct, Ubi-C1300-Bsr-d1 (abbreviated as Bsrd1ox). Each of the pANDA-Bsr-d1RNAi and Ubi-C1300-Bsr-d1 constructs was introduced into TP309 through Agrobacterium-mediated transformation as described previously (Li et al., 2016). The regenerated transgenic plants carrying Bsr-d1RNAi or Bsr-d1ox were selected with hygromycin. PCR-based genotyping was performed to verify the presence of the transgene as previously described (Chen et al., 2010). Overexpression and RNAi of Bsr-d1 in the transgenic lines was confirmed by qRT-PCR. The UbiC1300-Os10 g39170 overexpression construct (abbreviated as Os10 g39170ox) was generated following the same method as above. For CRISPR (Clustered regularly interspaced short palindromic repeats)/Cas9 construction, codon optimized hSpCas9 (Cong et al., 2013) was linked to the maize ubiquitin promoter (UBI) in an intermediate plasmid, and then this expression cassette were insert into binary pCAMBIA1300 (Cambia, Australia) which contains the HPT (hygromycin B phosphotransferase) gene. The original BsaI site present in the pCAMBIA1300 backbone was removed using point mutation kit (Transgen, China). A fragment, containing a OsU6 promoter (Feng et al., 2013), a negative selection marker gene ccdB flanked by two BsaI sites and a sgRNA derived from pX260 (Cong et al., 2013), was insert into this vector using In-fusion cloning kit (Takara, Japan) to produce the CRISPR/Cas9 binary vector pBGK032. E. coli strain DB3.1 was used for maintaining this binary vector. The 23 bp targeting sequences (including PAM) were selected within the target genes and their targeting specificity was confirmed using a Blast search against the rice genome (https://blast.ncbi.nlm.nih.gov/Blast.cgi) (Hsu et al., 2013). The designed targeting sequences were synthesized and annealed to form the oligo adaptors. Vector pBGK032 was digested by BsaI and purified using DNA purification kit (Tiangen, China). A ligation reaction (10 mL) containing 10 ng of the digested pBGK032 vector and 0.05 mM oligo adaptor was carried out and directly transformed to E. coli competent cells to produce CRISPR/Cas9 plasmids. The CRISPR/Cas9 plasmids were introduced into A. tumefaciens strain EHA105. Transformation of rice was performed as described previously (Nishimura et al., 2006). Genomic DNA was extracted from these transformants and primer pairs flanking the designed target site were used for PCR amplification. The PCR products (300-500 bp) were sequenced. QUANTIFICATION AND STATISTICAL ANALYSIS All quantification analyses on lesion area and H2O2 area were conducted in Photoshop. The statistical analyses were performed using SPSS 13.0. All values are presented with mean ± SEM and the number (n) of samples is indicated in the legend. Statistically significant differences between control and experimental groups were determined by one-way ANOVA, with p values < 0.01.

Cell 170, 114–126.e1–e5, June 29, 2017 e5

Supplemental Figures

Figure S1. Blast Resistance of Digu and LTH under Field Condition and Sequence Comparison of Bsr-d1 between Digu and LTH, Related to Figure 1 (A) Picture of the representative plants of Digu and LTH from natural rice blast field. Enlarged pictures of Digu (a) and LTH (b) are shown on the right. Bar = 40 cm. (B) Quantitation of blast resistance of Digu and LTH. Photographs of two representative leaves (left panel) and quantitation of lesion area (right panel) of each variety are shown. Lesion area was calculated based on pixels taken with Photoshop. Formula: lesion area per leaf/leaf area = pixel of lesion area per leaf/pixel of leaf area. Data are represented as mean ± SEM. (C) Comparison of Digu and LTH Bsr-d1 CDS sequences. CDSs of Bsr-d1 are identical between Digu and LTH. (D) Promoter sequences of Bsr-d1 from Digu and LTH. Identical bases are highlighted in gray.

(legend on next page)

Figure S2. Elucidation of the Role of the bsr-d1 Allele in Blast Disease Resistance, Related to Figures 1 and 2 (A) Bsr-d1 RNA expression in RILs. Analysis was performed on five representative RILs with the Digu-type SNP33G and five representative RILs with the LTH-type SNP33-A in the presence or absence of M. oryzae infection. (B and C) Real time RT-PCR analyses were performed on three Digu-type SNP33G (B) and five LTH-type SNP33-A (C) rice accessions 24 hr post inoculation with or without (mock) M. oryzae isolate ZB15. Rice Digu and LTH were included. (D) Bsr-d1 expression was specifically reduced in the Bsr-d1RNAi transgenic plants. Expression of LOC_Os03 g32230 (Bsr-d1) was determined in Bsr-d1RNAi lines Bsr-d1RNAi#1 and Bsr-d1RNAi#2 by using qRT-PCR. (E) Alignment the CDSs of LOC_Os03 g32230 (Bsr-d1) and LOC_Os03 g32220. Identical bases are highlighted in black. (F) Expression of LOC_Os03 g32220 (high sequence similarity to Bsr-d1) was determined in Bsr-d1RNAi lines Bsr-d1RNAi#1 and Bsr-d1RNAi#2 by using qRT-PCR. (G) Determination of the blast resistance on segregants derived from the Bsr-d1RNAi#1 transgenic line. The photographs of representative leaves post punchinoculation (left panel) and measurement of Bsr-d1 expression (right panel) from the segregants, A1 through A9, and the control (CK) TP309 are shown. (H) Determination of the blast resistance on segregants derived from the Bsr-d1RNAi#2 transgenic line. The photographs of leaves post inoculation (left panel) and measurement of Bsr-d1 expression (right panel) from the segregants, B1 through B9, and TP309 are shown similarly as in (F). In both (F) and (G), PCR-based genotyping with the primer pair specific for the hygromycin (Hyg) gene was performed to determine whether the plants contained (represented by ‘+’) or lacked (represented by ‘-’) the transgene Bsr-d1RNAi. Blast isolate ZB15 was used for inoculations (conidial concentration of 5 3 105/mL). (I) Determination of Bsr-d1 RNA expression in the Bsr-d1 overexpression plants. Leaf samples collected from Bsr-d1 overexpression lines Bsr-d1ox#1 and Bsrd1ox#2, and wild-type TP309 at the seedling stage were subjected to RNA extraction followed by qRT-PCR analysis. In (A), (B), (C), (D), (F), (G), (H) and (I), qRTPCR results are normalized with the Ubq5 reference gene. Error bars represent SEMs from three replicates. Similar results were obtained from three independent biological experiments. Asterisks represent significant differences (*p < 0.01).

Figure S3. Determination on the Resistance of the Bsr-d1 Knockout Transgenic Lines, Related to Figures 2 and 3 (A) Two separated target sites designed for knocking out the Bsr-d1 gene by CRISPR/Cas9 system. (B) Verification of the knockout (KO) lines by PCR-based sequencing. Two representative transgenic lines (abbreviated as Bsr-d1KO#1 and Bsr-d1KO#2, respectively) are generated from TP309 genetic background. (C) Picture of two leaves from two independent Bsr-d1KO transgenic lines, Bsr-d1KO#1 and Bsr-d1KO#2, and the control TP309. (D) Statistical analysis on lesion lengths on the inoculated leaves. Blast isolates Zhong10-8-14-eGFP, ZE-1, NC24, B04, NC10, HN41, 9920-2, Guy11, and FJ089-1 were each used for inoculation with a conidial concentration of 5 3 105 /mL. Data are represented as mean ± SEM. Asterisks represent significant differences (*p < 0.01). Three independent biological experiments were performed with similar results obtained.

Figure S4. Characterization of DNA-Binding Site in BSR-D1 and the Target Genes, Related to Figure 4 (A) Amino acid sequence of BSR-D1. Conserved C2H2 domains are underlined. The conserved amino acids of zinc finger, ‘‘C’’ and ‘‘H,’’ are highlighted with black and in large size. (B) A binding motif identified in the overlapping BSR-D1 binding peaks. The motif is found in the promoters of two peroxidase genes based on ChIP-seq. (C) BSR-D1 binding profile in the promoters of two peroxidase genes. Red rectangle represents peak summits. Numbers represent the positions of bases. (D) The target site designed for knocking out the Os10 g39170 gene by CRISPR/Cas9 system. (E) Verification of the knockout lines by PCR-based sequencing. Two representative transgenic lines (abbreviated as Os10 g39170KO#1 and Os10 g39170KO#2, respectively) for Os10 g39170 knockout are generated from TP309 genetic background. (F) Determination of Os10 g39170 RNA expression in the Os10 g39170 overexpression plants. Leaf samples collected from Os10 g39170 overexpression lines Os10 g39170ox#1 and Os10 g39170ox#2, and wild-type TP309 at the seedling stage were subjected to RNA extraction followed by qRT-PCR analysis. qRT-PCR results are normalized with the Ubq5 reference gene. Error bars represent SEMs from three replicates. Similar results were obtained from three independent biological experiments. Data are represented as mean ± SEM. Asterisks represent significant differences (*p < 0.01).

Figure S5. Identification and Function Determination of MYB Transcription Factors that Regulate Bsr-d1 and BSR-D1-Target Genes, Related to Figures 5 and 6 (A) Screening of five MYB transcription factors (TFs) based on expression patterns. The expression pattern of each of the five MYB TFs was significantly corelated with that of Bsr-d1 (p % 0.01). Expression patterns of five MYB TFs (left panel) and their correlation coefficiences with Bsr-d1 expression (right panel) are shown. (B) Schematic of the chimeric genes. ADH1-GAL4AD-MYB is used as the effector construct. Bsr-d1-miniHis3-HIS3 is used as the reporter construct. (C) Screening of MYB TFs bind to the Bsr-d1 promoter. Comparing to the control, the growth of the yeast co-transformed with Os01 g34060-Rec2 and Bsr-d1 promoter-pHIS2 is obviously inhibited in SD/-Trp-Leu-His medium with 30% 3-AT in the yeast one-hybrid assay, indicating that MYBS1 binds to the Bsr-d1 promoter and therefore inhibits its transcription. However, yeast cells co-transformed with other four MYB TFs with Bsr-d1 promoter-pHIS2 grow as well in SD/-Trp-Leu-His medium with 30% 3-AT as the control. The yeast co-transformed with the reporter pGADT7-Rec2 and Bsr-d1 promoter-pHIS2 is used as control. (D) Two separated target sites designed for knocking out the Mybs1 gene by CRISPR/Cas9 system. (E) Verification of the knockout lines by PCR-based sequencing. Two representative transgenic lines (abbreviated as Mybs1-KO#1 and Mybs1-KO#2, respectively) for Mybs1 knockout are generated from TP309 genetic background. (F) RNA expression of Bsr-d1 in the Mybs1 knockout transgenic plants. Samples were collected from Mybs1-KO#1 and Mybs1-KO#2, and TP309 at the seedling stage. (G) Expression levels of two Bsr-d1-regulated genes in Mybs1 knockout plants. qRT-PCR was performed for Mybs1-KO#1, Mybs1-KO#2, and TP309. RNA was prepared from leaf samples at the four-leaf stage. In (F) and (G), qRT-PCR results are normalized with the Ubq5 reference gene. Error bars represent the SEM of three replicates. Asterisks represent significant differences (*p < 0.01). Similar results were obtained from three independent biological experiments.

(legend continued on next page)

(H) Blast resistance determination of rice plants to ensure the inoculation is effective. The leaves of Digu and Lijiangxintuanheigu (LTH) before and post inoculation using spraying (upper panel) and punch inoculation (lower panel) methods were used for Mybs1 transcriptional expression analyses as shown in (L) and (M). Pictures were taken seven days post inoculation. Blast isolate ZB15 was used for inoculations. Blast-105 represent the conidial concentration of 5 3 105 /mL, and Blast-107 represent the conidial concentration of 5 3 107 /mL. (I) Mybs1 expression levels in Digu and LTH post mock- or M. oryzae-inoculation using spraying method. (J) Mybs1 expression levels in Digu and LTH post mock- or M. oryzae-inoculation using punch inoculation. In (L) and (M), the Mybs1 expression levels are normalized with the Ubq5 reference gene. RNA was prepared from leaf samples at the four-leaf stage. Error bars represent the SEMs from three replicates. Similar results were obtained from three independent biological experiments.

Figure S6. The bsr-d1 Is a Novel Allele in Disease Resistance and Possesses the Potential Value in Application, Related to Figures 1 and 2 (A) RNA levels of the OsNAC111 and OsERF922 genes in Bsr-d1 knockout and TP309 plants. (B) RNA levels of the OsRbohA, OsRbohB and OsRbohE genes in the Bsr-d1 knockout and TP309 plants. In (A) and (B), qRT-PCR results are normalized with the Ubq5 reference gene. RNA was prepared from leaf samples at the four-leaf stage. Error bars represent the SEM of three replicates. Similar results were obtained from three independent biological experiments.

(legend continued on next page)

(C) Agronomic traits of Bsr-d1KO lines and TP309. Samples were collected from reprehensive Bsr-d1KO lines Bsr-d1KO#1 and Bsr-d1KO#2, and wild-type TP309. Agronomic traits including plant height, tiller number per plant, seed setting rate, 1000-seed-weight and grain weight per plant are collected from a total of 16 plants. (D) Mature seeds of Bsr-d1KO and TP309. (E) Percentage of grains with chalkiness with three replicates. (F) Enlarged view of mature seeds in white light background. (G) Degree of endosperm chalkiness of TP309 and Bsr-d1KO plants with three replicates. (H) Geographic distribution of rice varieties containing the Digu-type allele carrying SNP33-G. Red dots represent the countries where rice varieties with the Digutype SNP33 are cultivated, and yellow areas represent rice-growing areas. The genome sequences of 3,000 accessions (Li et al., 2014) are subjected to SNP analysis through CallSNP method. Three hundred and thirteen accessions homozygous for Digu-type SNP33-G from 26 countries are identified.

Figure S7. The Schematic of SNP Calling Pipelines, Related to Figure 1