Skip to main content
NCBI home page
Genetics logoLink to Genetics
. 2022 Feb 10;220(4):iyac019. doi: 10.1093/genetics/iyac019

Fitness benefits play a vital role in the retention of the Pi-ta susceptible alleles

Jia Liu 1,#, Suobing Zhang 2,#, Pengfei Xie 1, Long Wang 3, Jia-Yu Xue 1,4, Yanmei Zhang 1, Ruisen Lu 1, Yueyu Hang 1, Yue Wang 1,✉, Xiaoqin Sun 1,✉
Editor: A Paterson
PMCID: PMC8982021  PMID: 35143673

Abstract

In plants, large numbers of R genes, which segregate as loci with alternative alleles conferring different levels of disease resistance to pathogens, have been maintained over a long period of evolution. The reason why hosts harbor susceptible alleles in view of their null contribution to resistance is unclear. In rice, a single copy gene, Pi-ta, segregates for 2 expressed clades of alleles, 1 resistant and the other susceptible. We simulated loss-of-function of the Pi-ta susceptible allele using the CRISPR/Cas9 system to detect subsequent fitness changes and obtained insights into fitness effects related to the retention of the Pi-ta susceptible allele. Our creation of an artificial knockout of the Pi-ta susceptible allele suffered fitness-related trait declines of up to 49% in terms of filled grain yield upon the loss of Pi-ta function. The Pi-ta susceptible alleles might serve as an off-switch to downstream immune signaling, thus contributing to the fine-tuning of plant defense responses. The results demonstrated that the susceptible Pi-ta alleles should have evolved pleiotropic functions, facilitating their retention in populations. As Pi-ta is a single copy gene with no paralogs in the genome, its function cannot be compensated by an alternative gene; whereas most other R genes form gene clusters by tandem duplications, and the function could be compensated by paralogs with high sequence similarity. This attempt to evaluate the fitness effects of the R gene in crops indicates that not all disease resistance genes incur fitness costs, which also provides a plausible explanation for how host genomes can tolerate the possible genetic load associated with a vast repertoire of R genes.

Keywords: fitness, Pi-ta, the susceptible alleles, knockout

Introduction

In plants, defense pathways against microbial and fungal pathogens are triggered by recognition and signal transduction systems encoded by large, diverse multigene families, called R genes. Although genetic resistance provides an effective method for the control of plant diseases, mounting a defensive response frequently comes with the cost of reductions in growth and reproduction, carrying critical implications for natural and agricultural plant populations (Brown 2002). Research on fitness effects was pioneered by Vander Plank (1963) in his study of the resistance of potato to late blight (Phytophthora infestans) and has been documented in other plants. For example, the wheat (Triticum aestivum) streak mosaic virus R gene Wsm1 is associated with a mean yield reduction of 21% (Sharp et al. 2002), the wheat stem rust R gene Sr26 has a 9% yield penalty (Brown 2002), and the barley (Hordem vulgare) mloR gene has a 4.2% yield penalty (Jørgensen 1992). Research on the fitness effects of resistance is currently undergoing a surge of interest, especially the study of the molecular mechanisms underlying the trade-offs between yield and disease resistance, and researchers are proposing new breeding strategies for selecting high-yield cultivars with minimum fitness costs in crops. For example, Pigm and Xa4 have been found to confer disease resistance to the rice blast fungus (Magnaporthe oryzae) and the rice bacterial blight (Xanthomonas oryzae pv. oryzae), respectively, both without compromising grain yield (Deng et al. 2017; Hu et al. 2017).

R genes can evolve rapidly, presumably to keep pace with the evolution of pathogens in order to retain their ability to identify and respond to rapidly evolving pathogens. Molecular genetic studies of plant R genes have revealed extensive variation in R gene loci, with fitness effects acting as one of the key selective drivers of R gene evolution. Mathematical models generally suggest that a trade-off between the costs of resistance in pathogen-free environments and the benefits of resistance under infection is required to explain the long-lived presence/absence polymorphisms of several R genes (Bergelson et al. 2001). Indeed, using isolines, a high cost of resistance measured in the absence of disease has been determined for 2 R genes, Rps5 and Rpm1, in Arabidopsis thaliana (Tian et al. 2003; Karasov et al. 2014). Both Rps5 and Rpm1 exist in nature as long-lived, presence/absence polymorphisms for the entire gene for resistance (R) and susceptibility (S), respectively. In both cases, resistant isolines suffer a 5–10% fitness cost relative to null isolines in the absence of disease. Interestingly, our previous fitness trials testing the consequences of the Rps2 gene (Macqueen et al. 2016), which exists as an ancient balanced polymorphism with 2 long-lived clades of alleles,1 resistant and the other susceptible to Pseudomonas syringae pv. avrRpt2, found no fitness cost for encoding the resistance allele in the absence of infection. An a priori hypothesis is that loci that exhibit presence/absence polymorphisms are the most likely to exact large fitness costs, while those loci that exhibit non-presence/absence polymorphisms may have negligible costs. Such a hypothesis seeks to explain how host genomes can tolerate the possible genetic load associated with a vast repertoire of R genes.

Rice blast disease is one of the most severe diseases threatening rice production worldwide. The Pi-ta gene in rice has been effectively used to fight against blast disease globally (Bryan et al. 2000; Jia et al. 2000). Pi-ta is a single copy gene that encodes a predicted cytoplasmic protein with a nucleotide binding site (NBS) and leucine-rich repeats (LRRs) (Bryan et al. 2000). A single amino acid substitution, serine (Ser) to alanine (Ala) at the position of 918 in the LRR of the Pi-ta protein, revealed the direct interaction with AVR-Pita and the resistance specificity to blast pathogen M. oryazae (Bryan et al. 2000; Jia et al. 2000). Interestingly, Pi-ta is present in every rice accession studied to date, none of which have missing data or deletions called for Pi-ta (Jia et al. 2016). The undeniable benefits of resistance should have driven the resistant alleles to fixation, while in theory there would be no reason for hosts to harbor susceptible alleles. It is therefore unclear why rice populations still retain disease-susceptible individuals along with disease-resistant individuals. In fact, even though function of the resistant Pi-ta alleles to the blast disease is well known, the function of the susceptible Pi-ta alleles remains poorly known. We hypothesized that the Pi-ta susceptible allele must have another beneficial function to permit its retention. Wang et al. (2015) reported that the Pi-ta susceptible alleles were associated with heavier seed weight based on genome-wide association analysis (GWAS), which was indicative of the fitness effect of the susceptible alleles on yield. However, it is still unclear if the effect of Pi-ta on yield is due to a direct effect of the gene itself or another yield-related gene located in the same genomic region, as genes that are linked to one R gene may also affect yield (Brown 2002).

Empirical evidence for the fitness effects of R genes is still controversial because some R genes may be tightly linked to other yield-related genes (Ortelli et al. 1996), and fitness effects are affected by environmental conditions and may vary by evaluation method (Laine 2016). Therefore, pairs of transgenic lines with rigorously controlled backgrounds and field trials should be the best materials for studying the fitness effects of individual R genes.

Herein, to discover the role of fitness effects on the retention of the Pi-ta susceptible allele, we simulated a loss of function of the Pi-ta susceptible allele to detect any subsequent fitness changes. Loss of function of the susceptible Pi-ta susceptible allele was induced by targeted mutation using the CRISPR/Cas9 system. We screened a rigorously controlled line without mutations in siblings of the knockout mutants as the control. We measured and compared the fitness in isogenic lines with and without a functional Pi-ta allele in the absence of disease. We additionally verified the robustness of our results by using 5 independent isogenic lines for the mutation of Pi-ta. Our results revealed that the knockout of the Pi-ta susceptible alleles suffered drastic fitness-related trait declines (up to 49%) in terms of filled grain yields. Gene expression profiling of these isogenic lines showed that the Pi-ta susceptible alleles might contribute to the fine-tuning of the downstream defense response, explaining their retention in the genome. This is a novel attempt to evaluate the fitness effect of the R gene in a rigorously controlled genetic background, especially in crops. Compared to typical model plants, the balance between immunity and yield in crops would draw more attention from breeders and researchers. In theory, this study will shed light on why and how plants can afford such a huge repertoire of susceptible R gene alleles in the genome. This phenomenon may occur because some of these susceptible alleles have fitness benefits, and they are not null alleles at all. In practice, this study will provide support to the concept that not all disease resistance genes will incur fitness costs, as universally acknowledged.

Materials and methods

Plant material and growth

The background of transgenic plants was Nipponbare (Oryza sativa L. ssp. japonica). All rice plants were grown in a greenhouse at our institute at 28–35°C or in fields in our experimental station under normal growth conditions in Nanjing, Jiangsu province, China. The experimental station is specialized for genetically modified crop planting as permitted by the Chinese Ministry of Agriculture.

Vector construction

The pTGE1 vector was constructed according to the method described by Xie and Yang (2013). Briefly, the primers g11S/g11AS covering target site sequence 11 (Fig. 1a and Supplementary File 1) were synthesized by Genscript (www.genscript.com.cn, February 9, 2022) and combined by annealing. The corresponding primers were cloned into the sgRNA expression cassette of pGREB31 (Addgene). The pTGE2 vector was constructed as described above based on target site sequence 12 (Fig. 1b and Supplementary File 1).

Fig. 1.

Fig. 1.

Open in a new tab

CRISPR/Cas9-induced Pi-ta gene modification in rice. a) Pi-ta target loci. The gRNA11 and gRNA12 targeting sites were designed in the first and second exons of Pi-ta. The target sites are labeled in black lowercase letters. The protospacer adjacent motif (PAM) sequences are underlined. b) The vector used to transform rice. gRNA11 and gRNA12 were assembled into the pGREB31 expression vector for Nipponbare transformation. c) Nucleotide sequences at the target site in the 6 T0 mutant rice plants. The target site nucleotides are indicated by black capital letters and black dashes. The PAM site is underlined. The red dashes indicate the deleted nucleotides. The red capital letters indicate the inserted nucleotides. The numbers on the right indicate the number of nucleotides involved. “−” and “+” indicate the deletion and insertion of the indicated number of nucleotides, respectively.

Rice transformation

The Cas9/sgRNA-expressing binary vectors (pTGE1 and pTGE2) were transformed into the Agrobacterium tumefaciens strain LBA4404 by electroporation. Agrobacterium-mediated transformation of embryogenic calli derived from “Nipponbare” was performed according to procedures detailed in Hiei et al. (1994). Briefly, hygromycin-containing medium was used to select hygromycin-resistant calli, and then the hygromycin-resistant calli were transferred to regeneration medium for the regeneration of transgenic plants. After 2–3 months of cultivation, transgenic seedlings were transferred to the greenhouse until maturity.

Genetic and phenotypic characterization of isogenic lines

Genomic DNA was extracted from individual transgenic plants using SDS extraction (Dellaporta et al. 1983). All transgenic hygromycin-resistant T0 plants were characterized by PCR using the Cas9-specific primers Cas9S/Cas9AS (Fig. 1 and Supplementary File 1). Subsequently, all PCR-positive plants were subjected to PCR using the gene-specific primer pairs Pi-ta787/Pi-ta1900 and Pi-ta4678/Pi-ta5522 (Fig. 1 and Supplementary File 1) to amplify DNA fragments across the 2 target sites. The resulting PCR amplicons were then directly sequenced. The sequencing chromatograms with superimposed peaks of biallelic and heterozygous mutations were decoded by the degenerate sequence decoding (DSD) method-based web tool DSDecode (http://skl.scau.edu.cn/dsdecode/, February 9, 2022) (Liu et al. 2015).

The presence of CRISPR/Cas9 DNA and marker genes in gene-edited plants may cause adverse effects, such as an increased risk of off-target changes and may trigger regulation concerns when these plants are used in crop breeding. Here, we tried to screen for individuals free of transferred DNA (T-DNA) in the T1 generation. The T-DNA-free lines were screened by amplifying the genomic DNA of T1 generation lines using Cas9 gene-specific primers (Cas9S/Cas9AS) (Supplementary File 1). The T-DNA free lines were selected for sequencing of the target regions. T-DNA-free lines with homozygous or no mutations were selected for further study. T-DNA-free lines with homozygous mutations were treated as the mutant lines, and T-DNA-free lines with no mutations were treated as control.

For evaluation of blast disease, seedlings of isogenic lines (∼10 plants of each line) at the 3- to 4-leaf stage were spray-inoculated with isolate 2014-30-2C15, with spore suspensions of 10−5 spores mL−1 in a growth chamber. Magnaporthe oryzae strain 2014-30-2C15, containing the avirulence gene AVR-Pita, was kindly provided by Prof. Zuhua He, CAS Center for Excellence in Molecular Plant Sciences/Institute of Plant Physiology and Ecology. Disease evaluation was performed 7 days after spraying, and the lesion types on leaves were observed and scored on a scale of 0–5 (Silué et al. 2011).

DAB (3, 3′-diaminobenzidine) solution was used to evaluate H2O2 accumulation in isogenic lines. Staining was performed as previously described (Qiao et al. 2010).

Field and greenhouse fitness experiments

To reveal the evolutionarily relevant fitness effects of resistance, fitness traits were assessed under natural growing conditions and in the absence of avirulent M. oryzae. Seedlings of each of 7 Pi-ta lines were grown in the field under normal growth conditions in Nanjing. The experimental plot was partitioned into 20 blocks in a randomized block design, in which each block contained 4 replicates for each Pi-ta line. Each block consisted of 4 rows with 7 plants per row, with 1 meter (m) between rows and 0.9 m spacing between plants within rows. The field was hand-weeded once, and plants received no other protection from competition or pests. Roughly 12% of plants died, but these were evenly distributed among the lines.

The greenhouse fitness experiment was performed in 20 square buckets. Each bucket contained 4 replicates from each Pi-ta line in a randomized design. Plants were set out in 4 rows in each bucket, spaced by 0.3 m within rows and by 0.4 m between rows. Roughly 20% of plants died, but these were evenly distributed among the lines.

Four fitness proxies were measured in the field and the greenhouse fitness experiments: dry weight, height, number of filled grains, and filled grain yield.

All 7 Pi-ta lines were grown in sterile conditions. The sterile experiments included 10 seedlings per line. Seeds were surface-sterilized by brief vortexing in 75% ethanol and then soaked in 30% (v/v) sodium hypochlorite for 5 min and thoroughly washed. Seeds were germinated on ½MS media with agar in sterile beakers, and 7 seeds from different lines were evenly distributed in 1 beaker in a randomized design. Two-week-old seedlings were harvested to measure plant height, fresh weight, and dry weight.

Sampling for pathogens

A key step to accurately measure the costs of resistance was to confirm that no avirulent M. oryzae that would have interfered with the experiment were present at the field site.

Ninety-six plant samples representing all 7 lines were surface-sterilized and plated on PAD medium on days 30 and 60 of growth during the field and greenhouse experiments. Out of hundreds of colonies, 42 potentially distinct fungus types were identified by general appearance. Each type was sequenced for ITS (Khang et al. 2008) and compared to published sequences in GenBank. None of the fungus isolates were M. oryzae according to their ITS identity. Finally, we used PCR to screen for the presence of AVR-Pita. We screened with 4 combinations of primer pairs (Supplementary File 1) based on the published sequence, but AVR-Pita was not detected in any of our colonies.

Quantification of salicylic acid (SA), jasmonic acid (JA), and 1-aminocyclopropane-1-carboxylic acid (ACC)

Three replicates of X0 and X2 lines were grown in the greenhouse and harvested on Day 60 of growth.

JA was extracted and quantified as described by Schweizer et al. (1997). Briefly, 0.5–1.0 g of fresh leaves were used for JA purification and GC/MS analysis (Finnigan Trace GC–MS, USA), and 9,10-dihydro-JA was added as an internal standard.

SA was extracted from 1.0 g of fresh leaves and quantified with a high-performance liquid chromatography system equipped with fluorescence detection (LC-2010 AHT, Japan) as described by Metwally et al. (2003). Authentic SA was used for calibration.

Next, following the procedures described by Lizada and Yang (1979), ACC was extracted and quantified. One gram of fresh leaves was used for ACC purification and high-performance liquid chromatography/mass spectrometry analysis. Authentic ACC was used for calibration.

Whole transcriptome profiling

Three biological replicates (10 plants each) of each line were sampled for RNA sequencing. Total RNA was extracted using TRIzol reagent (Life Technologies). The qualified RNA samples were then used for library construction following the specifications of the TruSeq RNA Sample Preparation v2 Guide (Illumina, USA), and RNA sequencing was conducted on an Illumina Hiseq 2500 at Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China). We used SeqPrep to strip adaptors and/or merge paired reads that overlapped into single reads and used Sickle to remove low-quality reads. The clean data were then mapped to the reference genome of rice using HISAT2 v2.1.0. FPKM (fragments per kilobase per million mapped reads) was then calculated to estimate the expression levels of genes. DESeq2 v1.6.3 was used to analyze the differential gene expression between 2 samples, and genes with q < 0.05 and |log2_ratio| > 1 were identified as differentially expressed genes (DEGs). Gene ontology (GO; http://geneontology.org/, February 9, 2022) enrichment of the DEGs was determined by hypergeometric tests, in which the P-value was calculated and adjusted as the q-value. GO terms with q < 0.05 were considered significantly enriched. For pathway analysis, we mapped all the DEGs in terms of the Kyoto Encyclopedia of Genes and Genomes (KEGG) and retrieved the significantly enriched pathway with P-value ≤ 0.05 (Kanehisa and Goto 2000). The MapMan package from the Max Planck Institute of Molecular Plant Physiology, Germany was employed to obtain the graphical representation of the biotic stress response by DEGs (Thimm et al. 2004).

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis

Leaves were sampled from 2-month-old plants grown in the greenhouse under standard conditions. Total RNA isolation, first-strand cDNA synthesis, and qRT-PCR analysis were performed as described by Sharma et al. (2013). Three PCR replicates were run for each line. Experiments were conducted with 3 biological replicates for each sample, and 3 technical replicates were analyzed for each biological replicate. The ΔΔCT calculation method was used to calculate the relative expression level of each gene. For normalizing the relative mRNA level of individual gene in various RNA samples, EF-257 was used as an internal control gene (Yang et al. 2018). The list of primer sequences used for qRT-PCR analysis is provided in Supplementary File 1.

Statistical analysis

Data analysis was performed in Microsoft Excel and is presented as the mean ± standard error (SE) of biological replicates. Statistical analysis was performed using one-way analysis of variance (ANOVA); this was followed by Duncan’s test using SPSS (version 19.0) for data statistics. Different letters are used (P< 0.05) to present statistically significant differences, and identical letters are considered as statistically nonsignificant.

Sequence data analysis

Genome sequences and annotation files of 32 accessions of O. sativa and the related species Oryza brachyantha were downloaded from http://ricerc.sicau.edu.cn/RiceRC/ (February 9, 2022) and https://ftp.ensemblgenomes.org/pub/plants/release-51/fasta/triticum_aestivum/dna/ (February 9, 2022), respectively. Pi-ta gene identification was performed on the annotated protein as described previously (Bryan et al. 2000).

DNA sequences were aligned using ClustalW, and phylogenetic analysis was performed using the Maximum Likelihood method with MEGA X software (Kumar et al. 2018). DnaSP 6 software (Rozas et al. 2017) was used to calculate the values of π (nucleotide diversity), θ (theta from S, theta-W), Ka/Ks (Ka = the rate of nonsynonymous substitution, Ks = the rate of synonymous substitution), and D (Tajima’s D) (Tajima 1989).

Results

Generation of isogenic lines of the Pi-ta gene using CRISPR/Cas9

Pi-ta is a single copy gene that encodes a predicted cytoplasmic protein with NBS and LRR domains (Bryan et al. 2000; Jia et al. 2000). Nipponbare (O. sativa L. ssp. japonica), which serves as the genetic background in this study, possesses a susceptible allele of the Pi-ta gene in its genome. Two target sites for CRISPR/Cas9 editing were located in proximity to the 5′ end and the 3′ end of the gene (Fig. 1a; 6–25 and 2,436–2,455 bp in the open reading frame, respectively). Synthesized oligos were inserted into the CRISPR/Cas9 binary vector for editing (Fig. 1b). Subsequently, the 2 constructed vectors were transformed into Nipponbare using the Agrobacterium-mediated method.

In T0 transgenic plants, we screened 5 lines with targeted mutations (Fig. 1c), 4 of which (X1, X3, X4, and X6) showed homozygous mutations, while only 1 (X2’) was heterozygous. A total of 4 mutations were achieved at Target 11, in contrast to the 1 mutation found at Target 12. All mutations were located in front of the amino acid residue 918, which was reported to be responsible for gene-for-gene specificity to rice blast (Bryan et al. 2000). Due to the frameshift that resulted from the mutations, the deduced amino acid sequences of Pi-ta were 58 (X3 and X4), 59 (X1), and 832 (X6) amino acid residues compared to the 928 amino acid residues in Nipponbare.

Mutants of T0 were planted and 253 T1 plants were screened to analyze their genetic transformation patterns. T-DNA free mutants were selected using Cas9 gene-specific primers (Cas9S/Cas9AS, Supplementary File 1). The results showed that 42 mutants were not amplified to the Cas9 vector sequence. These mutants were thus termed as T-DNA-free plants. Of the T-DNA-free plants, the inheritance of mutations at the target sites of lines X1, X3, X4, and X6 was further checked by sequencing. In X2’, which had heterozygous mutations at the target site, homozygous individuals and nonmutated individuals were screened in the segregated T1 siblings, termed as X2 and X0, respectively. In order to create rigorously controlled isogenic lines with the same genetic background, X0 was treated as a control without mutations (Pi-ta0) as it underwent the same cultivation, transformation, and regeneration procedures as the other 5 Pi-ta knockout mutant lines (collectively, Pi-ta-). Nipponbare was included as the wild type in this study and was termed X7 (Pi-taWT). All the lines were then selfed to the T3 generation for further analysis.

The CRISPR-GE online tool was used to identify potential off-target sites for the 2 Cas9 sgRNAs from the edited plants. No sgRNAs had predicted off-target sites with fewer than 2-nt mismatches, and only one off-target site was predicted with 7-nt mismatches for sgRNA11, suggesting that the designed spacers were highly specific. For this one potential off-target site, PCR amplification followed by Sanger sequencing was used, and it was found that no mutations were induced in any edited lines (Supplementary File 1). To further assess possible off-target effects, 2 (X2 and X6) of 5 edited lines were arbitrarily selected for whole genome sequencing and the mutations were identified using a pipeline as described in a previous study (Wang et al. 2019). The targeted mutations were confirmed in the sequenced plants. No novel mutations were found in the other ∼500 R genes identified in both rice genomes, and no common mutations were shared by X2 and X6.

To explore the influence of mutations in the Pi-ta gene on disease resistance, all 7 isogenic lines, as well as 9311 as the resistant control, were inoculated with M. oryzae strain 2014-30-2C15 carrying AVR-Pita. Interestingly, 5 Pi-ta knockout mutant lines (X1, X2, X3, X4, and X6) relieved the susceptibility to blast strain 2014-30-2C15, with smaller disease lesions observed on the leaves than in both the wild type (Nipponbare) and line X0, without mutations (Supplementary File 2, a and b). Furthermore, DAB staining was used as an indicator of H2O2 accumulation; intense brown staining was observed in the leaves of the 5 Pi-ta knockout mutant lines (X1, X2, X3, X4, and X6) together with X0, but such a signal was not detected in the wild-type line (X7) (Supplementary File 2c).

Fitness-related traits decline in Pi-ta knockout mutants in the absence of avirulent pathogens

To test the fitness effect of the Pi-ta gene in the absence of infection, we measured the relative fitness of Pi-ta-/Pi-ta0/Pi-taWT isolines (representing knockout, nonmutated, and wild-type backgrounds, respectively) in field and greenhouse experiments in the absence of AVR-Pita.

In the field experiment, in 19 out of 20 comparisons using 4 fitness proxies for 5 independent knockout mutants, Pi-ta− isogenic lines demonstrated significantly lower performance than Pi-taWT (Fig. 2, a–d and Supplementary File 3). In terms of dry weight, height, number of filled grains, and filled grain yield, Pi-ta− isogenic lines suffered up to 41%, 28%, 42%, and 49% reductions, respectively, relative to the Pi-taWT line (Fig. 2, a–d and Supplementary File 3).

Fig. 2.

Fig. 2.

Open in a new tab

Significant fitness variations among isogenic lines in the absence of avirulent pathogens. X1, X2, X3, X4, and X6 are independent knockout mutant lines of the Pi-ta susceptible allele; X7 is the wild-type background line; X0 is the nonmutated line screened in the X2 siblings, treated as the control without mutations in Pi-ta. Lines with different independent mutations in the Pi-ta gene were tested in the field, in greenhouse experiments, and in sterile conditions. Bars with different letters are significantly different. Statistical differences among the agronomic traits were detected by Duncan’s multiple range test (P < 0.05). a–d) Field fitness results. a) Dry weight. b) Height. c) Number of filled grains. d) Filled grain yield. e–h) Greenhouse fitness results. e) Dry weight. f) Height. g) Number of filled grains. h) Filled grain yield. i, j) Sterile condition fitness results. i) Shoot fresh weight of seedling. j) Shoot length of seedling. Values are means ± standard error (± SE).

In 18 of 20 comparisons using 4 fitness proxies for 5 independent knockout mutants, Pi-ta− isogenic lines demonstrated significantly lower performance than Pi-ta0 (Fig. 2, a–d and Supplementary File 3). In terms of dry weight, height, number of filled grains, and filled grain yield, Pi-ta− isogenic lines suffered up to 38%, 19%, 35%, and 41% reductions, respectively, relative to the Pi-ta0 line (Fig. 2, a–d and Supplementary File 3).

Taken together, these results demonstrate the drastic decline of fitness-related traits in Pi-ta knockout mutants in the absence of avirulent pathogens, which suggests that normal functioning of the Pi-ta susceptible allele in Nipponbare is beneficial in the absence of known Pi-ta-mediated pathogens carrying AVR-Pita.

To verify the robustness of our results, this fitness experiment was repeated in a greenhouse that was known to be free of Pi-ta-recognized pathogens. Similar to the field experiment, Pi-ta− isogenic lines exhibited significantly lower performance in terms of dry weight, height, number of filled grains, and filled grain yield relative to both Pi-ta0 and Pi-taWT isolines, in 10 out of 20 comparisons between mutant lines and Pi-ta0; and in 20 out of 20 comparisons between mutant lines and Pi-taWT (Fig. 2, e–h and Supplementary File 4). Thus, our greenhouse results recapitulated the results observed in the field. Interestingly, we also observed performance variations between field and greenhouse experiments. As proposed by Macqueen et al. (2016), resource abundance in the environment influences the costs of resistance. In poor environments, the cost of resistance is reduced; that is, if the individuals perform poorly enough, then it may become more difficult to distinguish the performance of resistant and susceptible individuals. For rice in the current study, the field provided sufficient resources to the plants, whereas in the greenhouse, the square buckets had limited resources. Thus, the plants in the greenhouse tended to survive rather than to initiate immunity, consequently homogenizing the performance of the mutant lines and the control. This might in part account for the nonsignificance in certain comparisons between the mutant lines and controls in the greenhouse experiment.

One possibility to explain the observed fitness-related trait decline of all Pi-ta knockout mutants in the absence of M. oryzae (AVR-Pita) is that the presence of a different and undetected pathogen recognized by the allele of Pi-ta in Pi-ta0 or Pi-taWT isolines in the field may have provided a benefit to isolines carrying the normally functioning Pi-ta allele.

As a final confirmation that the observed fitness-related trait decline was not due to an interaction with an unknown pathogen, we grew our isogenic lines in sterile conditions on agar (Supplementary File 5). Again, Pi-ta0 and Pi-taWT isolines had a higher weight and height than Pi-ta− plants at 14 days (Fig. 2, i and j and Supplementary File 6). This result excluded the possibility that the presence of Pi-ta carried a fitness benefit because of the recognition of pathogens.

Collectively, these results demonstrate a beneficial function of the susceptible allele of Pi-ta in the absence of avirulent pathogens. In addition, the results did not reveal significant variations in fitness between lines with different targeted mutation sites.

Pi-ta-associated changes in defense response gene expression in the absence of avirulent pathogens

Expression levels can alter the penetrance of phenotypes, and alternation of the expression of Pi-ta can lead to nonspecific activation of the hypersensitive response (HR) or even lethality, if expression levels are too high (Wang et al. 2019). To elucidate the molecular components involved in the fitness-related trait decline in Pi-ta knockout mutants, we first evaluated the expression of the Pi-ta gene by qRT-PCR in the Pi-ta− (X1, X2, X3, X4, and X6), Pi-taWT (X7), and Pi-ta0 (X0) lines in the absence of avirulent pathogens (Fig. 3). In many cases, Pi-ta expression was 1.2- to 2.0-fold higher in these isolines than in the wild-type line, with the exception of X6. Expression of the Pi-ta gene was significantly lower in the X6 line compared with wild-type plants, suggesting considerable variation in the expression of the Pi-ta gene across targeted mutation sites.

Fig. 3.

Fig. 3.

Open in a new tab

The expression of various PRs in leaves of 2-month-old plants grown in the greenhouse under standard conditions. Gene expression in each line has been shown relative to that in X0. Data are means ± SE. *, **, and *** indicate significant differences determined by the t-test at P < 0.05, 0.01, and 0.001, respectively; ns, not significant.

The 5 Pi-ta− (X1, X2, X3, X4, and X6) lines showed substantial accumulation of H2O2 and relieved susceptibility to the blast strain, which indicated that defense responses could have been switched on due to the Pi-ta mutation. To verify this assumption, qRT-PCR analyses of some pathogen-related genes (PRs) and defense genes were further conducted. The relative expression level of oxidative stress-related protein gene POX22.3 (peroxidase) was highly up-regulated in all Pi-ta− lines (Fig. 3), which was consistent with the DAB staining results. The secondary metabolite pathway gene OsPAL4 (phenylalanine ammonialyase) was also up-regulated about 1.39–2.14-fold in the Pi-ta− lines (Fig. 3). The pathogen-induced systemic acquired resistance (SAR) marker gene NH1 (OsNPR1) was up-regulated about 1.61–4.29-fold in the Pi-ta− lines (Fig. 3), indicating that SAR was possibly activated in the mutant. In addition, most of the PR genes, PR1a, PR1b, JIOsPR10, and PBZ1, were induced in the Pi-ta− lines. PR1a, PR1b, JIOsPR10, and PBZ1 were up-regulated to high levels in the Pi-ta− lines (Fig. 3). Compared with Pi-ta0, JIOsPR10 was up-regulated about 4.09–18.71-fold in the Pi-ta− lines (Fig. 3). PBZ1 was up-regulated about 5.39–7.48-fold in the Pi-ta− lines (Fig. 3). The relative expression levels of PR1a and PR1b were up to about 20-fold and about 70-fold, respectively, in the Pi-ta− lines (Fig. 3). This indicates that extensive defense responses might be activated in the Pi-ta− lines. Moreover, the activation of PRs and other defense genes was always associated with the occurrence of weakness in different independent mutant lines.

The 2 other genes, Ptr and Pi-42, were then selected as the Pi-ta-related genes, and their expression patterns were detected. Ptr, which encodes an atypical protein with an armadillo repeat domain, is required for Pi-ta resistance (Zhao et al. 2018). Pi-42 was reported to function together with the Pi-ta gene as a sensor/helper pair (Wang et al., 2019). Ptr expression was induced by about 2.01–3.10-fold in the Pi-ta− lines, while Pi-42 did not exhibit significantly differential expression between Pi-ta− and Pi-ta0 (Fig. 3).

Moreover, we compared the expression profiles of 2 isolines with rigorously controlled backgrounds, X2 and X0, where only 1 nucleotide was inserted in the coding region of X2. The X2 line had 50 genes that were up-regulated and 50 genes that were down-regulated relative to the X0 line (Fig. 4a). Enrichment of GO terms revealed that 31 biological process-related GO terms were significantly enriched in a comparison between X2 and X0 (Fig. 4b). Genes up-regulated in X2 plants were mainly enriched for GO annotations involved in the integrin-mediated signaling pathway and mRNA cleavage (Supplementary File 7, P = 3.6 × 10 −3, 1.76 × 10 −2), whereas genes down-regulated in X2 plants were enriched for inositol phosphate-mediated signaling, virus response, and hormone biosynthetic process annotations (Supplementary File 7, P = 3.6 × 10−3, 1.41 × 10−2, 2.3 × 10−2).

Fig. 4.

Fig. 4.

Open in a new tab

Pi-ta knockout mutant line X2 differentially expresses signal transduction, mRNA cleavage, and hormone biosynthesis genes relative to its sibling line (X0) without mutation in Pi-ta. X2 is the knockout mutant line of the Pi-ta susceptible allele; X0 is the nonmutated line screened in the X2 siblings, treated as the control without mutations in Pi-ta. a) Heatmaps and dendrograms of DEG sets. Genes are in rows, and biological replicates are in columns, with both dendrograms grouped by similarity of expression in the gene set displayed. b) MapMan analysis illustrating the DEGs involved in biotic stress. DEGs with |Log2FC|>1 and P < 0.05 were imported into the MapMan tool. Up and down-regulated are represented in blue and red, respectively.

KEGG annotation revealed that DEGs were significantly enriched in 2 pathways: zeatin biosynthesis and phenylpropanoid biosynthesis. In the zeatin biosynthesis pathway, CYP735A4, a major player in regulating cytokinin concentrations and hence influence plant growth and development (Tsai et al. 2012), was up-regulated in X2, whereas OsUGT710C2, which is reported to be an immune-suppressive factor (Kenney et al. 2020), was down-regulated in X2, demonstrating relaxed suppression of the downstream immune response. Phenylpropanoid biosynthesis seems to be important in rice in response to M. oryzae (Zhang et al. 2019). In this pathway, 2 DEGs, including Os3bglu7 and OsCCR, which were documented to play a role in defense-related processes in rice (Park et al. 2017), were unregulated in X2.

Plant hormones, including JA, SA, and ethylene (E), are all key regulators of defense responses. The enrichment of hormone biosynthetic genes in the X2 line prompted us to examine JA, SA, and ACC as the precursors of E levels in both the X2 and X0 lines. As shown in Fig. 5, JA, SA, and ACC concentrations were significantly higher in X2 plants than in X0 plants (with fold changes ranging from 1.43 to 1.72, P< 0.05 by ANOVA). These data indicate that loss of function of the Pi-ta gene induced by 1 nucleotide insertion in the coding region caused major alterations in the endogenous levels of JA, SA, and ACC.

Fig. 5.

Fig. 5.

Open in a new tab

Effect of Pi-ta knockout on accumulation of JA, SA, and ACC. X2 is the knockout mutant line of the Pi-ta susceptible allele; X0 is the nonmutated line screened in the X2 siblings, treated as the control without mutations in Pi-ta. Three replicates of the X0 and X2 lines were grown in the greenhouse and harvested on days 60 of growth for determination of JA (a), SA (b), and 1-ACC (c). Data are means ± SE. Bars with different letters are significantly different. Statistical differences among the relative hormone levels were detected by Duncan’s multiple range test (P < 0.05).

Evolutionary pattern of Pi-ta and 2 related genes

We identified orthologs of Pi-ta in 32 well-assembled rice genomes (Qin et al. 2021) and found that Pi-ta was present in each accession, including the close relative O. brachyantha, indicating an ancestral origin of the Pi-ta gene. Previous studies have reported that Pi-ta allelic sequences fall into 2 types: the 918-Ala-type (resistant) and 918-Ser-type (susceptible). Therefore, 32 rice accessions were divided into 2 types, with 10 resistant-type sequences and 22 susceptible-type sequences. Given the distinct differentiation of the 2 Pi-ta types, we further surveyed genetic variations between the 2 types. The 10 resistant-type sequences harbored no nucleotide polymorphisms. Among the 22 susceptible-type sequences, the nucleotide diversity value π was 0.00120, and the θ value from segregating sites was 0.00166, indicating low diversity based on previously published criteria (Yang et al. 2008). Therefore, the Pi-ta susceptible group may have experienced a low rate of evolution, comparable to another R gene, Rps4 of Arabidopsis (Bergelson et al. 2001). To examine the evolutionary dynamics of the Pi-ta gene, natural selection was evaluated by Tajima’s D test. Interestingly, for the susceptible group, the Tajima’s D value was −0.04134, which was significantly deviated (P< 0.05) from neutrality. Furthermore, the nonsynonymous to synonymous substitution ratio (Ka/Ks) for the susceptible group was determined to be 0.7438 (<1). These results suggested that the Pi-ta susceptible group may have experienced recurrent selective sweeps, in congruence with conclusions in Huang et al. (2008), which was conducive to decreasing the nucleotide variations of the susceptible alleles within the population and, therefore, maintaining the likelihood of functions other than recognizing pathogens.

Plant disease resistance genes often function in pairs or complex networks to mediate innate immunity to pathogens (Wang et al. 2019). Two genes were reported to function together with Pi-ta, including Pi-42, which is known as the helper gene of Pi-ta, and Ptr, which is required for Pi-ta resistance (Zhao et al. 2018; Wang et al. 2019). We further surveyed the genetic variations of Pi-42, Ptr, and Pi-ta to elucidate whether there were evolutionary constraints of Pi-42 and Ptr on Pi-ta, while accounting for the retention of the Pi-ta alleles in populations. We investigated the association between Pi-ta and the 2 other genes by analyzing differentiation [an Fst estimator based on nucleotide diversities, as reported in Hudson et al. (1992)] between the 2 Pi-ta types. The genetic differentiation between all types was significant (Fst = 0.71878, 0.47319, 0.33782; P= 0.0062, 0.0004, 0.0379 for Pi-ta, Ptr, and Pi-42, respectively). Thus, the sequence variations of Ptr and Pi-42 were correlated with genetic differentiation of Pi-ta. This is in congruence with the phylogenetic trees of these 3 genes (Supplementary File 8), which share similar topologies with each other. Although evolutionary analysis revealed tight relationships between the 3 genes, how they interact remains largely unknown.

Discussion

Drastic fitness-related trait declines in Pi-ta knockout mutants might be induced by autoimmunity or pleiotropy

Pi-ta is one of the most well-characterized blast R genes of O. sativa and induces defense responses against the blast fungus carrying the avirulence gene AVR-Pita (Bryan et al. 2000). It is located near the centromere of rice chromosome 12. Pi-ta encodes a predicted CC-NBS-LRR-type protein. The physical interaction between Pi-ta and AVR-Pita was dependent on a single amino acid at the 918th codon position located in the LRD of Pi-ta. Further DNA sequence analysis revealed unusually low DNA polymorphism of the Pi-ta allele among rice cultivars (Jia et al. 2016). In this study, the knockout mutants of the Pi-ta susceptible allele exhibited a significant decline in plant dry weight, height, number of filled grains, and filled grain yield relative to both the wild-type and nonmutated individuals. Gene expression profiling and elevated endogenous hormones of the isogenic lines suggested that the Pi-ta susceptible alleles might serve as an off-switch to downstream immune signaling, contributing to the fine-tuning of plant defense responses. Thus, the susceptible alleles should have the potential to evolved pleiotropic functions, facilitating their retention in populations. The evolution of R genes is characterized by a history of gene duplications, expanding from only a few ancestral single genes into a large gene family containing approximately 500 members in rice. Some members show high sequence similarity, thus can serve as an alternative of functional compensation for each other (Deng et al. 2017). But this is not the case for the Pi-ta gene. Pi-ta originated extremely early, which could date back to an early stage in Oryza evolution (Huang et al. 2008). Notably, it remains as a single copy gene without expansion in all Oryza species. Therefore, its function could not be compensated by other R genes once Pi-ta is lost. Furthermore, it’s likely that only when the susceptible alleles harbor pleoitropic, beneficial effects should they be retained by selection.

In our previous research, Pi-ta knockout in another japonica cultivar, Wuyungeng24, also led to deleterious phenotypic effects, including shorter and weaker plants as well as frequent deaths (Wang et al. 2019). Moreover, initial attempts to knockout the Pi-ta resistant allele in Tetep also failed due to nonviable progeny. Therefore, loss of function of the Pi-ta gene is always associated with the occurrence of weakness.

First, overstimulation or mis-regulation of the immune system in rice induced by the loss of function of the Pi-ta gene is the most plausible explanation for the drastic fitness-related trait decline in Pi-ta knockout mutants. In the absence of pathogens, plant immunity is under tight control to avoid the activation of defense responses, as failure to do so can lead to autoimmunity that compromises plant growth and development. More generally, functional R gene alleles, when expressed, have the potential to incur a physiological cost because of expression or mis-expression. Many autoimmune mutants have been reported, most of which are associated with dwarfism and often spontaneous cell death (Rodriguez et al. 2016).

Plant immune receptors are generally maintained in inactive states and are activated only upon the detection of pathogens. The benefit of carrying a functional allele of Pi-ta appears to result from its function as a negative regulator of the defense response, as the loss of function of Pi-ta causes the induction of a number of genes involved in response to stimuli, signaling transduction, defensive barriers, defense-related metabolic biosynthesis, and elevated hormone concentrations, together with increased H2O2 accumulation, eventually resulting in enhanced resistance to rice blast. Pi-ta has been reported to form interacting pairs with Pi-42, in which the Pi-42 gene (known as the helper) activates defense signaling upon sensing the presence of a blast pathogen, while Pi-ta (known as the sensor) recognizes pathogen effectors and acts as an inhibitor on the helper to prevent autoimmunity in the absence of the pathogen (Wang et al. 2019). Due to the frameshift that resulted from the targeted mutations, the Pi-ta genes in the X1, X2, X3, and X4 lines had a much shorter truncated protein, ranging from 58 (X3 and X4) to 59 (X1 and X2) amino acid residues, thus likely losing repression to its helper gene Pi-42 and downstream defense signaling. However, qRT-PCR experiments revealed no significant changes in the expression levels of paired Pi-42 in the mutant lines, suggesting that this process was likely initiated at the post-transcriptional level. In the X6 line, the targeted mutation in the C-terminal of the Pi-ta gene led to a truncated LRR domain. For Rps5, another CNL disease resistance gene, its LRR domain was reported to function to suppress gene activation in the absence of relevant pathogens (Qi et al. 2012). Therefore, the truncated LRR domain in the X6 line might be insufficient for the inhibition of Pi-ta autoactivation. These findings thus suggest that the activation of PRs and other defense genes most likely occur due to downstream defense signaling initiated by helper NLR proteins after losing their sensor suppressors, or even by the Pi-ta gene itself.

Second, the critical function of Pi-ta might provide a clear explanation for the fitness reduction in the mutant lines. Pi-ta may have pleiotropic functions other than recognizing pathogens. In addition to the immune receptor function, some R genes are involved in signaling cascades important for additional cellular processes, such as drought tolerance, development, and photomorphogenesis (Tameling and Joosten 2007). One example yields a clue as to the linkage between low nucleotide polymorphisms and functional pleiotropy. Rps4 of Arabidopsis not only has conferred resistance to P. syringae carrying the effector AvrRPS4 (Gassmann et al. 1999), but is also involved in phyB signaling (Faigón-Soverna et al. 2006). A relative lack of polymorphisms was also found in Rps4; only a single amino acid polymorphism was detected in its LRR region between resistant and susceptible alleles, which was the same as in Pi-ta (Bergelson et al. 2001). Pi-ta has undergone recurrent selective sweeps, unlike the diversifying selection of most other R genes. Selection is conducive to reducing the genetic diversity of the Pi-ta gene within populations and, therefore, maintaining the likelihood of functions other than recognizing pathogens. The loss of function of Pi-ta changes the expression of a number of genes involved in not only defense response but also processes including the photorespiratory pathway, leaf development, and auxin biosynthetic process (Supplementary File 8). Therefore, we suggest that Pi-ta might have the potential to evolve other physiological functions in addition to resistance. Thus far, rice germplasm without a Pi-ta homolog has not been identified; further indicating that the Pi-ta gene may play physiological roles in addition to resistance to M. oryzae. Our previous research has also demonstrated that a beneficial pleiotropic function of Rps2 in Arabidopsis is measurable in the absence of pathogens (Macqueen et al. 2016).

Last, the dramatic decline in fitness-related traits in the mutant lines may have resulted from mutations induced by regeneration from cell culture or off-target by the use of the CRISPR/Cas9 system. There is always the risk that transgenic rice plants will carry background mutations due to somaclonal variation. In rice, after a 5-month cell culture, mutations were introduced at the rate of 1.74 single nucleotide polymorphisms/Mb, at least 248-fold higher than the spontaneous mutation rate (Miyao et al. 2012). To eliminate the effects of background mutations on fitness, all the lines in this study were selfed to the T3 generation for further analysis. Moreover, 2 (X2 and X6) of 5 edited lines were arbitrarily selected for whole-genome sequencing, and the mutations were identified. No novel mutations were found in the ∼500 R genes identified in the rice genome, and no common mutations were shared by X2 and X6, indicating that the decline in fitness-related traits might not result from mutations in other R genes or unknown common functional genes. In addition, plants showed consistent fitness-related trait declines among different independent mutant lines, demonstrating a causal link between the loss of function of the Pi-ta gene and the decline of fitness-related traits. Nevertheless, it is impossible to rule out the likely effects of background mutations, as they may indeed occur, caused by regeneration, or off-target of CRISPR/Cas9, or even spontaneously. Actually, the spontaneous mutation rate was estimated to be 3.2 × 10−9 per site per meiosis in rice (Yang et al. 2015). The background effect could also be the reason that the pairs of X1 and X2, as well as X3 and X4, shared the same mutations in the Pi-ta gene but exhibited variations in their expression levels and agricultural traits.

Fitness benefits play a vital role in the retention of Pi-ta susceptible alleles

In plants, a large number of R genes that segregate as loci with alternative alleles conferring different levels of disease resistance to pathogens have been maintained over a long period of evolution (Michelmore and Meyers 1998). As such, there would be no reason for hosts to harbor susceptible alleles in view of their lack of contribution to resistance. The question of why rice populations still support disease-susceptible individuals of the Pi-ta gene along with disease-resistant individuals is an interesting research topic. There are several possible explanations for the long-term maintenance of these susceptible alternatives.

First, rice germplasm without a Pi-ta homolog has not been identified, suggesting that the Pi-ta gene might play physiological roles in addition to resistance to M. oryzae (Jia et al. 2016). Furthermore, the discovery of potential multiple Pi-ta products from alternative splicing and exon skipping raises the possibility that they may play a role in recognizing diverse pathogen signals (Costanzo and Jia 2009).

Our creation of an artificial knockout of the Pi-ta susceptible allele in Nipponbare revealed a fitness-related trait decline of up to 49% upon the loss of Pi-ta’s function. The benefit of carrying a functional allele of Pi-ta appears to result from its function as a negative regulator of the defense response. Serving as an off-switch to its helper Pi-42 and downstream immune signaling, Pi-ta was shown to be involved in the fine-tuning of the disease resistance response, as demonstrated in a study on the alternative alleles of Rps2 (Macqueen et al. 2016). The critical function of Pi-ta provides a clear explanation for why the susceptible alleles have not been eliminated, as they typically are for Rpm1 and Rps5. This result may help explain why the Pi-ta locus does not harbor true susceptible deletion mutants.

In plants, duplicated gene copies are assumed to persist as reservoirs for functionally distinct pathogen recognition alleles, providing sources for generating novel specificities by mutation and/or intergenic recombination (Michelmore and Meyers, 1998). In theory, single-copy genes could also restore new alleles through mutation or recombination between the diverse alleles inherited from parents in natural or manual hybrids. Multiple avirulent AVR-Pita haplotypes, determining the specificity of Pi-ta-mediated resistance, have been identified in commercial rice fields (Jia et al. 2016). To keep pace with the rapid evolution of the pathogens, more resistant Pi-ta alleles are urgently needed. Spontaneous mutations or the reshuffling of variations in current alleles enables the continuous generation of diverse, novel alleles. The only resistant Pi-ta haplotype detected thus far was inferred to derive from an ancestral haplotype H2 in wild rice (Huang et al. 2008). Up to 100 haplotypes of the Pi-ta gene discovered in Oryza species might serve as an enlarged reservoir of variations to combat various pathogens. Furthermore, plants are continuously exposed to the threat of diverse pathogenic microbes, including viral, bacterial, or fungal pathogens, in their natural habitats. It is possible that there may be an unknown microbe that interacts with the Pi-ta susceptible alleles, and function constraints consequently enabling the maintenance and diversification of these alleles in populations.

Finally, unlike in wild plants, the yield, quality, and agronomic properties, such as optimal height and heading time, of arable crops are all normally considered more important than disease resistance (Brown 2002). The need to prioritize other traits over disease resistance may have an impact on plant breeding programs. If resistance to a disease is not an important commercial target, if the susceptible alleles provide fitness benefits such as yield increases, and if breeders therefore select for it, the frequency of the susceptible alleles in breeder germplasms will increase as a result of artificial selection. Thus, in over 10,000 years of domestication of cultivated rice, plant breeding can be regarded as an essential force in the retention of the susceptible Pi-ta alleles.

Our results were consistent with the fitness benefits of Rps2 in Arabidopsis (Macqueen et al. 2016). Both the present study and the study in Arabidopsis show that the R gene alleles, including the susceptible one, contribute to the regulation of gene expression and thus present a pleiotropic effect, explaining their maintenance in the genome.

Experimental studies to evaluate the fitness effects of R genes, as described above, have focused on loci with very simple genetic architectures. R genes show a high degree of copy number variation within and across plant species, and many of them are organized in tandem arrays (Michelmore and Meyers 1998). How plants with a vast repertoire of R genes coordinate the fitness costs and benefits of the alternative alleles will surely be a topic worthy of further research.

Data availability

Strains and plasmids are available upon request. The RNA-seq data that support the findings of this study are available in the Dryad repository (https://doi.org/10.5061/dryad.6hdr7sr1c; Sun et al. 2021).

Supplemental material is available at GENETICS online.

Supplementary Material

iyac019_Supplementary_Data
Click here for additional data file. (2.3MB, docx)

Acknowledgments

The authors thank Dacheng Tian for the donation of the rice material. They also thank Zuhua He for kindly providing M. oryzae strains.

Funding

This work was supported by grants from the National Natural Science Foundation of China to X.Q.S. (Grant no. 31470448; 32171483), the Jiangsu Key Laboratory of Plant Resources Research and Utilization grant to X.Q.S. (JSPKLB201921), and the Changshu Agricultural Production and Public Service Project and Natural Science Foundation of Jiangsu Province to J.L. (BK20200289).

Conflicts of interest

The authors declare that there is no conflict of interest.

Literature cited

  1. Bergelson J, Kreitman M, Stahl EA, Tian DC.. Evolutionary dynamics of plant R-genes. Science. 2001;292(5525):2281–2285. [DOI] [PubMed] [Google Scholar]
  2. Brown JKM. Yield penalties of disease resistance in crops. Curr Opin Plant Biol. 2002;5(4):339–344. [DOI] [PubMed] [Google Scholar]
  3. Bryan GT, Wu KS, Farrall L, Jia Y, Hershey HP, McAdams SA, Faulk KN, Donaldson GK, Tarchini R, Valent B, et al. A single amino acid difference distinguishes resistant and susceptible alleles of the rice blast resistance gene Pi-ta. Plant Cell. 2000;12(11):2033–2045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Costanzo S, Jia YL.. Alternatively spliced transcripts of Pi-ta blast resistance gene in Oryza sativa. Plant Sci. 2009;177(5):468–478. [Google Scholar]
  5. Dellaporta SL, Wood J, Hicks JB.. A plant DNA minipreparation: version II. Plant Mol Biol Rep. 1983;1(4):19–21. [Google Scholar]
  6. Deng Y, Zhai K, Xie Z, Yang D, Zhu X, Liu J, Wang X, Qin P, Yang Y, Zhang G, et al. Epigenetic regulation of antagonistic receptors confers rice blast resistance with yield balance. Science. 2017;355(6328):962–965. [DOI] [PubMed] [Google Scholar]
  7. Faigón-Soverna A, Harmon FG, Storani L, Karayekov E, Staneloni RJ, Gassmann W, Más P, Casal JJ, Kay SA, Yanovsky MJ, et al. A constitutive shade-avoidance mutant implicates TIR-NBS-LRR proteins in Arabidopsis photomorphogenic development. Plant Cell. 2006;18(11):2919–2928. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Gassmann W, Hinsch ME, Staskawicz BJ.. The Arabidopsis RPS4 bacterial-resistance gene is a member of the TIR-NBS-LRR family of disease-resistance genes. Plant J. 1999;20(3):265–277. [DOI] [PubMed] [Google Scholar]
  9. Hiei Y, Ohta S, Komari T, Kumashiro T.. Efficient transformation of rice (Oryza sativa L.) mediated by Agrobacterium and sequence analysis of the boundaries of the T-DNA. Plant J. 1994;6(2):271–282. [DOI] [PubMed] [Google Scholar]
  10. Hu K, Cao J, Zhang J, Xia F, Ke Y, Zhang H, Xie W, Liu H, Cui Y, Cao Y, et al. Improvement of multiple agronomic traits by a disease resistance gene via cell wall reinforcement. Nat Plants. 2017;3:17009. [DOI] [PubMed] [Google Scholar]
  11. Huang CL, Hwang SY, Chiang YC, Lin T.. Molecular evolution of the Pi-ta gene resistant to rice blast in wild rice (Oryza rufipogon). Genetics. 2008;179(3):1527–1538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hudson RR, Slatkin M, Maddison WP.. Estimation of levels of gene flow from DNA sequence data. Genetics. 1992;132(2):583–589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Jia Y, Zhou E, Lee S, Bianco T.. Coevolutionary dynamics of rice blast resistance gene Pi-ta and Magnaporthe oryzae avirulence gene AVR-Pita 1. Phytopathology. 2016;106(7):676–683. [DOI] [PubMed] [Google Scholar]
  14. Jia YL, Mcadams SA, Bryan GT, Hershey HP, Valent B.. Direct interaction of resistance gene and avirulence gene products confers rice blast resistance. EMBO J. 2000;19(15):4004–4014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Jørgensen IH. Discovery, characterization and exploitation of MLO powdery mildew resistance in barley. Euphytica. 1992;63(1–2):141–152. [Google Scholar]
  16. Kanehisa M, Goto S.. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 2000;28(1):27–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Karasov TL, Kniskern JM, Gao L, DeYoung BJ, Ding J, Dubiella U, Lastra RO, Nallu S, Roux F, Innes RW, et al. The long-term maintenance of a resistance polymorphism through diffuse interactions. Nature. 2014;512(7515):436–440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kenney E, Yaparla A, Hawdon JM, O'Halloran DM, Grayfer L, Eleftherianos I.. A putative UDP-glycosyltransferase from Heterorhabditis bacteriophora suppresses antimicrobial peptide gene expression and factors related to ecdysone signaling. Sci Rep. 2020;10(1):12312. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Khang CH, Park SY, Lee YH, Valent B, Kang S.. Genome organization and evolution of the AVR-Pita avirulence gene family in the Magnaporthe grisea species complex. Mol Plant Microbe Interact. 2008;21(5):658–670. [DOI] [PubMed] [Google Scholar]
  20. Kumar S, Stecher G, Li M, Knyaz C, Tamura K.. MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol Biol Evol. 2018;35(6):1547–1549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Laine A. Disease resistance: not so costly after all. Nat Plants. 2016;2:16121. [DOI] [PubMed] [Google Scholar]
  22. Liu W, Xie X, Ma X, Li J, Chen J, Liu Y-G.. DSDecode: a web-based tool for decoding of sequencing chromatograms for genotyping of targeted mutations. Mol Plant. 2015;8(9):1431–1433. [DOI] [PubMed] [Google Scholar]
  23. Lizada C, Yang SF.. A simple and sensitive assay for 1-aminocyclopropane-1-carboxylic acid. Anal Biochem. 1979;100(1):140–145. [DOI] [PubMed] [Google Scholar]
  24. Macqueen A, Sun XQ, Bergelson J.. Genetic architecture and pleiotropy shape costs of Rps2-mediated resistance in Arabidopsis thaliana. Nat Plants. 2016;2:16110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Metwally A, Finkemeier I, Georgi M, Dietz K.. Salicylic acid alleviates the cadmium toxicity in barley seedlings. Plant Physiol. 2003;132(1):272–281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Michelmore RW, Meyers BC.. Clusters of resistance genes in plants evolve by divergent selection and a birth-and-death process. Genome Res. 1998;8(11):1113–1130. [DOI] [PubMed] [Google Scholar]
  27. Miyao A, Nakagome M, Ohnuma T, Yamagata H, Kanamori H, Katayose Y, Takahashi A, Matsumoto T, Hirochika H.. Molecular spectrum of somaclonal variation in regenerated rice revealed by whole-genome sequencing. Plant Cell Physiol. 2012;53(1):256–264. [DOI] [PubMed] [Google Scholar]
  28. Ortelli S, Winzeler H, Winzeler M, Fried PM, Nosberger J.. Leaf rust resistance gene Lr9 and winter wheat yield reduction: I. Yield and yield components. Crop Sci. 1996;36(6):1590–1595. [Google Scholar]
  29. Park HL, Bhoo SH, Kwon M, Lee SW, Cho MH.. Biochemical and expression analyses of the rice Cinnamoyl-CoA reductase gene family. Front Plant Sci. 2017;8:2099. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Qi D, DeYoung BJ, Roger WI.. Structure-function analysis of the coiled-coil and leucine-rich repeat domains of the RPS5 disease resistance protein. Plant Physiol. 2012;158(4):1819–1832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Qiao Y, Jiang W, Lee J, Park BSoo, Choi M‐S, Piao R, Woo M‐O, Roh J‐H, Han L, Paek N‐C, et al. 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. 2010;185(1):258–274. [DOI] [PubMed] [Google Scholar]
  32. Qin P, Lu H, Du H, Wang H, Chen W, Chen Z, He Q, Ou S, Zhang H, Li X, et al. Pan-genome analysis of 33 genetically diverse rice accessions reveals hidden genomic variations. Cell. 2021;184(13):3542–3558. [DOI] [PubMed] [Google Scholar]
  33. Rodriguez E, Ghoul HE, Mundy J, Petersen M.. Making sense of plant autoimmunity and ‘negative regulators’. FEBS J. 2016;283(8):1385–1391. [DOI] [PubMed] [Google Scholar]
  34. Rozas J, Ferrer-Mata A, Sánchez-DelBarrio JC, Guirao-Rico S, Librado P, Ramos-Onsins SE, Sánchez-Gracia A.. DnaSP 6: DNA sequence polymorphism analysis of large data sets. Mol Biol Evol. 2017;34(12):3299–3302. [DOI] [PubMed] [Google Scholar]
  35. Schweizer P, Buchala AJ, Metraux J.. Gene-expression patterns and levels of jasmonic acid in rice treated with the resistance inducer 2,6-dichloroisonicotinic acid. Plant Physiol. 1997;115(1):61–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sharma R, Priya P, Jain MK.. Modified expression of an auxin-responsive rice CC-type glutaredoxin gene affects multiple abiotic stress responses. Planta. 2013;238(5):871–884. [DOI] [PubMed] [Google Scholar]
  37. Sharp GL, Martin JM, Lanning SP, Blake NK, Brey CW, Sivamani E, Qu R, Talbert LE.. Field evaluation of transgenic and classical sources of wheat streak mosaic virus resistance. Crop Sci. 2002;42(1):105–110. [DOI] [PubMed] [Google Scholar]
  38. Silué D, Notteghem JL, Tharreau D.. Evidence of a gene-for-gene relationship in the Oryza sativa-magnaporthe grisea pathosystem. Phytopathology. 2011;82:577–580. [Google Scholar]
  39. Tajima F. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics. 1989;123(3):585–595. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tameling WIL, Joosten MHAJ.. The diverse roles of NB-LRR proteins in plants. Physiol Mol Plant Pathol. 2007;71(4–6):126–134. [Google Scholar]
  41. Thimm O, Bläsing O, Gibon Y, Nagel A, Meyer S, Krüger P, Selbig J, Müller LA, Rhee SY, Stitt M.. MAPMAN: a user-driven tool to display genomics data sets onto diagrams of metabolic pathways and other biological processes. Plant J. 2004;37(6):914–939. [DOI] [PubMed] [Google Scholar]
  42. Tian D, Traw MB, Chen J, Kreitman M, Bergelson J.. Fitness costs of R-gene-mediated resistance in Arabidopsis thaliana. Nature. 2003;423(6935):74–77. [DOI] [PubMed] [Google Scholar]
  43. Tsai Y-C, Weir NR, Hill K, Zhang W, Kim HJ, Shiu S-H, Schaller GE, Kieber JJ.. Characterization of genes involved in cytokinin signaling and metabolism from rice. Plant Physiol. 2012;158(4):1666–1684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Vander Plank JE. Plant Diseases: Epidemics and Control. New York and London: Academic Press; 1963. p. 1–349. [Google Scholar]
  45. Wang L, Zhao L, Zhang X, Zhang Q, Jia Y, Wang G, Li S, Tian D, Li W-H, Yang S.. Large-scale identification and functional analysis of NLR genes in blast resistance in the Tetep rice genome sequence. Proc Natl Acad Sci USA. 2019;116(37):18479–18487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Wang X, Jia MH, Ghai P, Lee FN, Jia YL.. Genome-wide association of rice blast disease resistance and yield-related components of rice. Mol Plant Microbe Interact. 2015;28(12):1383–1392. [DOI] [PubMed] [Google Scholar]
  47. Xie K, Yang Y.. RNA-guided genome editing in plants using a CRISPR–Cas system. Mol Plant. 2013;6(6):1975–1983. [DOI] [PubMed] [Google Scholar]
  48. Yang R, Li J, Zhang H, Yang F, Wu Z, Zhuo X, An X, Cheng Z, Zeng Q, Luo Q.. Transcriptome analysis and functional identification of Xa13 and Pi-ta orthologs in Oryza granulata. Plant Genome. 2018;11(3):170097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Yang S, Gu T, Pan C, Feng Z, Ding J, Hang Y, Chen J-Q, Tian D.. Genetic variation of NBS-LRR class resistance genes in rice lines. Theor Appl Genet. 2008;116(2):165–177. [DOI] [PubMed] [Google Scholar]
  50. Yang S, Wang L, Huang J, Zhang X, Yuan Y, Chen J-Q, Hurst LD, Tian D.. Parent-progeny sequencing indicates higher mutation rates in heterozygotes. Nature. 2015;523(7561):463–467. [DOI] [PubMed] [Google Scholar]
  51. Zhang W, Yan J, Li X, Xing Q, Chethana KWT, Zhao W.. Transcriptional response of grapevine to infection with the fungal pathogen Lasiodiplodia theobromae. Sci Rep. 2019;9(1):5387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Zhao H, Wang X, Jia Y, Minkenberg B, Wheatley M, Fan J, Jia MH, Famoso A, Edwards JD, Wamishe Y, et al. The rice blast resistance gene Ptr encodes an atypical protein required for broad-spectrum disease resistance. Nat Commun. 2018;9(1):2039. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

iyac019_Supplementary_Data
Click here for additional data file. (2.3MB, docx)

Data Availability Statement

Strains and plasmids are available upon request. The RNA-seq data that support the findings of this study are available in the Dryad repository (https://doi.org/10.5061/dryad.6hdr7sr1c; Sun et al. 2021).

Supplemental material is available at GENETICS online.

Articles from Genetics are provided here courtesy of Oxford University Press

RESOURCES