Recent Progress in Rice Broad-Spectrum Disease Resistance

Rice is one of the most important food crops in the world. However, stable rice production is constrained by various diseases, in particular rice blast, sheath blight, bacterial blight, and virus diseases. Breeding and cultivation of resistant rice varieties is the most effective method to control the infection of pathogens. Exploitation and utilization of the genetic determinants of broad-spectrum resistance represent a desired way to improve the resistance of susceptible rice varieties. Recently, researchers have focused on the identification of rice broad-spectrum disease resistance genes, which include R genes, defense-regulator genes, and quantitative trait loci (QTL) against two or more pathogen species or many isolates of the same pathogen species. The cloning of broad-spectrum disease resistance genes and understanding their underlying mechanisms not only provide new genetic resources for breeding broad-spectrum rice varieties, but also promote the development of new disease resistance breeding strategies, such as editing susceptibility and executor R genes. In this review, the most recent advances in the identification of broad-spectrum disease resistance genes in rice and their application in crop improvement through biotechnology approaches during the past 10 years are summarized.


Introduction
Rice (Oryza sativa L.) is the most important food crop, which is consumed by approximately 50% of the world's population, with its consumption growing dramatically in many parts of the world. Stable rice production is constrained by various biotic stresses, including fungal blast caused by Magnaporthe oryzae, sheath blight caused by Rhizoctonia solani, false smut caused by Ustilaginoidea virens, bakanae disease due to Fusarium fujikuroi, bacterial blight caused by Xanthomonas oryzae pv. oryzae (Xoo), bacterial leaf streak caused by Xanthomonas oryzae pv. oryzicola (Xoc), and virus diseases. The yield loss of rice caused by various diseases averages upward of 30%. Therefore, it is critical to adopt effective means to control these diseases to ensure global food security. In addition to encouraging farmers to exercise good farming practices, application of pesticides remains one of the main methods of disease control, but the increase in costs and their harmful effects on the environment and operators cannot be discounted. These make the farmers largely dependent on the cultivation of new resistant varieties, which is considered to be the most effective method so far.
Broad-spectrum resistance, which refers to resistance against two or more types of pathogen species or the majority of races/isolates of the same pathogen species, is one of the ultimate goals of breeders for rice improvement [1]. Exploitation of the genetic determinants of broad-spectrum resistance will improve the resistance of the susceptible varieties. On this account, breeders and biotechnologists are trying to obtain the source of broad-spectrum resistance to understand and utilize the genetics underlying this process.
With the development of rice molecular biology, functional genomics, and gene editing technology, great progress has been made in the broad-spectrum disease resistance genes in rice. It is worth mentioning that many extremely important broad-spectrum disease resistance genes and their mechanisms and applications were reported 10 years ago. These works have been well summarized in several reviews and will not be highlighted here [1,2]. This review focuses on the progress in the identification of broad-spectrum disease resistance genes in rice and their application in crop improvement during the past 10 years.

R Genes Confer Broad-Spectrum Disease Resistance in Rice
The ability of plants to defend themselves against microbes is specified by disease resistance (R) genes-mediated resistance and basal resistance. Upon recognition of an invading pathogen, R proteins, mostly from the nucleotide-binding leucine-rich repeat receptor (NLR) family, detect the secreted pathogen effectors to activate a multitude of responses that ultimately lead to resistance. These responses include Ca 2+ influx, ROS (reactive oxygen species) accumulation, mitogen-activated protein (MAP) kinase activation, defense hormone signaling activation, and upregulation of defense-related genes [3]. In recent years, important progress has been achieved in cloning and mechanical analysis of R genes. These identified R genes provide not only new genetic resources for breeding broad-spectrum rice varieties, but also new strategies to improve resistance.
In addition to rice blast, bacterial blight, caused by Xoo, is also a globally devastating rice disease. In rice, at least 46 genes that confer dominant or recessive host resistance to Xoo have been identified, of which more than 16 R genes were cloned [27]. Among them, Xa7, Xa23, Xa41(t), and Xa47(t) were cloned as broad-spectrum resistance R genes in the past 10 years [27] (Figure 1, Table 1). Xa7, which encodes a 113 aa unknown protein, is a dominant R gene that provides broad-spectrum and extremely durable resistance to Xoo. The transcription of Xa7 is specifically activated by the Xoo isolates with transcription activator-like effector (TALE) AvrXa7 or PthXo3 to act as an executor [27]. Another executor R gene, Xa23, which is induced by TALE AvrXa23, confers extremely broad-spectrum resistance to Xoo [28]. xa41(t), an allele of sugar transporter OsSWEET14, confers resistance to half of the tested Xoo isolates [29]. Similar to Xa23, Xa47(t) is from the wild rice Oryza rufipogon. Xa47(t), encoding a NLR protein, is highly resistant to all tested ten Xoo isolates [30]. In addition to these cloned R genes, a broad-spectrum bacterial blight resistance gene Xa33 from Oryza nivara has been finely mapped [31]. In addition to rice blast, bacterial blight, caused by Xoo, is also a globally devastating rice disease. In rice, at least 46 genes that confer dominant or recessive host resistance to Xoo have been identified, of which more than 16 R genes were cloned [27]. Among them, Xa7, Xa23, Xa41(t), and Xa47(t) were cloned as broad-spectrum resistance R genes in the past 10 years [27] (Figure 1, Table 1). Xa7, which encodes a 113 aa unknown protein, is a dominant R gene that provides broad-spectrum and extremely durable resistance to Xoo. The transcription of Xa7 is specifically activated by the Xoo isolates with transcription activator-like effector (TALE) AvrXa7 or PthXo3 to act as an executor [27]. Another executor R gene, Xa23, which is induced by TALE AvrXa23, confers extremely broad-spectrum resistance to Xoo [28]. xa41(t), an allele of sugar transporter OsSWEET14, confers resistance to half of the tested Xoo isolates [29]. Similar to Xa23, Xa47(t) is from the wild rice Oryza rufipogon. Xa47(t), encoding a NLR protein, is highly resistant to all tested ten Xoo isolates Unlike rice blast and bacterial blight, no R gene for serious diseases sheath blight and rice false smut has been identified [32,33]. Moreover, only one rice stripe virus (RVS) resistance gene, STV11, has been cloned [34]. The molecular mechanisms underlying ricevirus interaction remain poorly understood. Therefore, so far, there is no broad-spectrum resistance R gene for these diseases.
For broad-spectrum resistance R genes, it is worth noting that the R genes are tagged as broad-spectrum resistance genes because they can resist multiple isolates of one pathogen rather than two or more types of pathogen species. However, in the case of many isolates tested, it is unlikely for any R gene to be resistant to only one isolate. At present, there is no standard in terms of how many isolates or what proportion of isolates an R gene confers resistance to for it to be claimed as a broad-spectrum resistance R gene. Furthermore, although so many broad-spectrum resistance genes have been identified, the mechanism of these genes mediating broad-spectrum resistance to rice disease is not yet clear. In the broadspectrum resistance R gene Pi9 case, the corresponding Avirulence gene AvrPi9 exists widely in various M. oryzae isolates [35]. To determine whether other blast R genes are similar, the cloning of their corresponding Avirulence genes and analysis of their distribution in M. oryzae isolates will give some hints. Similar to R genes against blast, the resistance spectrum of R genes against Xoo may also be determined by the distribution of corresponding Avirulence genes, TALE in most of cases, in Xoo isolates. Actually, considering current knowledge, it is difficult to predict the mechanism underlying R protein-mediated broadspectrum resistance in addition to the wide distribution of corresponding Avirulence genes. The research progress of the interactions between R protein and effectors from pathogens will increase our understanding of R protein-mediated broad-spectrum resistance. NBS-LRR 10 Xoo isolates 11 [30] 1 The pathogens and number of isolates to which resistance is conferred by the gene in the first column. Magnaporthe oryzae, Mo; Xanthomonas oryzae pv. oryzae, Xoo.

Defense Regulator Genes Contribute to Broad-Spectrum Disease Resistance
Differing from R genes, defense regulator genes often confer partial resistance to a broad spectrum of pathogen isolates or various pathogens. In the past 10 years, there were at least 56 broad-spectrum resistant defense regulator genes identified which positively or negatively regulate the resistance to rice diseases ( Figure 1, listed in Table 2). The proteins encoded by these genes are transcriptional factors, kinases, peroxidases, E3 ubiquitin ligases, ferredoxin-dependent glutamate synthases, glutaredoxins, etc. In this review, we classify these broad-spectrum resistant defense regulator genes according to the types of pathogens they resist.   OsBON3 Copine protein 1 Mo isolate, 3 Xoo isolates, 1 Rs isolate 5 [87] In the past 10 years, several broad-spectrum resistant defense regulator genes against M. oryzae were identified. Through a genome-wide association study (GWAS), a natural allele of a C2H2-type transcription factor bsr-d1 was identified in rice that confers nonrace-specific resistance to blast. This allele causes a lower gene expression level, and then downregulates expression of three peroxidase-encoding genes, Os05g04470, Os10g39170, and Perox3, resulting in accumulation of H 2 O 2 and enhanced broad-spectrum resistance to M. oryzae [36,37]. In addition, an MYB transcription factor (OsMYB30) is also involved in bsr-d1-mediated broad-spectrum blast resistance by activating the lignin biosynthesis genes Os4CL3 and Os4CL5 to strengthen cell walls [38]. The other three transcription factors, OsNAC60, OsWRKY45, and RRM (RNA recognition motif) protein PIBP1 (PigmRinteracting and blast resistance protein 1), also contribute to broad-spectrum blast resistance in rice. OsNAC60, which is a target of Osa-miR164a, negatively regulates rice immunity against the blast fungus M. oryzae by decreasing programmed cell death, ion leakage, ROS accumulation, callose deposition, and defense-related gene expression [39]. OsWRKY45 mediates the blast resistance of CC-NB-LRR protein Pb1 [40]. PIBP1 specifically interacts with PigmR and other similar NLRs, and it functions as an unconventional transcription factor to activate the expression of OsWAK14 and OsPAL1 to trigger blast resistance [41]. In addition to transcription factors, the RING protein OsBBI1 with E3 ligase activity and light-harvesting complex II protein LHCB5 are also involved in broad-spectrum blast resistance. OsBBI1 confers broad-spectrum resistance against M. oryzae by increasing H 2 O 2 accumulation in cells and modifying the cell-wall defense [42]. Phosphorylation of LHCB5 enhances resistance to M. oryzae through the accumulation of ROS in the chloroplast [43].
Several broad-spectrum resistant defense-regulator genes against Xoo were also identified in rice in the past 10 years. In rice, several genes involved in receptor-mediated broad-spectrum resistance and systemic acquired resistance (SAR) likely contribute to broad-spectrum resistance to Xoo. The XA21-binding protein XB25, a plant-specific ankyrin repeat (PANK) protein, contributes to the accumulation of receptor XA21 and maintenance of XA21-mediated broad-spectrum resistance to Xoo [44]. The endoplasmic reticulum (ER) chaperone, luminal-binding protein 3 (BiP3) negatively regulates resistance mediated by rice XA3, a receptor that provides broad-spectrum resistance to Xoo [45]. Overexpression of OsNPR1 (non-expressor of pathogenesis-related genes 1), a master gene for SAR in rice, greatly enhances resistance to Xoo [46]. Moreover, the cysteine-rich-receptor-like kinases (OsCRK6 and OsCRK10) are required for OsNPR1-mediated immunity [47]. In addition to CRK6 and CRK10, some kinases have been identified as conferring broad-spectrum resistance to Xoo. For instance, overexpression of a constitutively activated form of calcium-dependent protein kinases OsCDPK1 confers Xoo resistance by affecting OsPR10a expression in rice [48]. OsILA1, a Raf-like MAPKKK, functions as a negative regulator and acts upstream of the OsMAPKK4-OsMAPK6 cascade against Xoo [49]. Unlike the type of genes mentioned above, lc7, encoding a mutant ferredoxin-dependent glutamate synthase 1 (Fd-GOGAT1), promotes ROS accumulation in the leaves and has high broad-spectrum resistance against seven Xoo strains [50].
In addition, several genes have been reported as conferring broad-spectrum resistance against multiple pathogens other than both M. oryzae and Xoo. Aldehyde dehydrogenase OsALDH2B1 has a moonlight function as a transcriptional regulator to regulate a diverse range of biological processes involving G protein, brassinolide, jasmonic acid, and salicylic acid signaling pathways. Loss of function of OsALDH2B1 greatly enhanced resistance to M. oryzae, Xoo, and Xoc [74]. Similarly, rice phenylalanine ammonia-lyase gene OsPAL4 is associated with resistance to M. oryzae, Xoo, and Xoc [75]. Heat-shock factor OsHsfB4d binds the promoter and regulates the expression of a small heat-shock protein gene OsHsp18.0-CI to be resistant against Xoo and Xoc [76,77]. Moreover, suppression of phytoalesin-deficient 4 OsPAD4 results in increased susceptibility to the Xoo and Xoc [78]. Rice glutaredoxin gene OsGRXS15 and a novel NPR1 homolog gene OsNH5N16 contribute to broad-spectrum resistance to Xoo and F. fujikuroi by regulating the expression of PR genes related to SAR [79,80]. Abscisic acid, stress, and ripening 2 (ASR2) contributes to broad-spectrum resistance against Xoo and R. solani by regulating the expression of a defense regulator gene Os2H16 [81,82]. In contrast, 14-3-3 protein (GF14e) negatively affects cell death and disease resistance to Xoo and R. solani [83]. WRKY transcription factor OsWRKY30 and 1-aminocyclopropane-1carboxylic acid synthase gene ACS2 positively regulate the resistance against M. oryzae and R. solani [84,85]. Moreover, methyl esterase-like gene OsMESL and copine genes OsBON1 and OsBON3 are critical suppressors of immunity to M. oryzae, Xoo, and R. solani [86,87].
These excellent studies on broad-spectrum resistant defense regulator genes show multiple characteristics. Firstly, with the increasing attention to broad-spectrum resistance, the reports of broad-spectrum resistance related genes have increased sharply in the past 10 years [1]. For breeders, whether these genes also show broad-spectrum disease resistance in the natural field environment is still the focus of attention. Secondly, the connections between these broad-spectrum resistant defense regulator genes and their relationships with R genes remain largely unclear due to limited experimental evidence. It is possible that these broad-spectrum resistance defense regulator genes function in the convergence point of the crosstalk between the pathways of basal and R protein-mediated resistances or between the pathways initiated by different R proteins [1]. Thirdly, only a few broad-spectrum resistant defense regulator genes mediate resistance with little or no yield penalties. The tradeoff between broad-spectrum resistant defense regulator genes and rice yield is one of the important limiting factors, as summarized in Chen's review [2]. Last but not least, there are rare examples of using these disease resistance-related genes to obtain broad-spectrum disease-resistant varieties in breeding programs. Although it was very difficult to effectively use these broad-spectrum resistant defense regulator genes in molecular breeding of rice until now, identification of natural variations/alleles of these genes from rice varieties with excellent agronomic traits, artificial mutation, and genome-editing technology would provide important methods for broad-spectrum disease resistance.

Identification of Broad-Spectrum Disease Resistance Loci by QTL Mapping and GWAS Analysis
Broad-spectrum resistance is a polygenic trait, whereby a combinatorial effect of major and minor genes mediates this trait [88]. With the advances of next-generation DNA sequencing and high-density molecular marker platforms, various quantitative trait loci (QTL) against rice blast, sheath blight, and/or bacterial leaf blight have been mapped to locate the source of these traits in the past 10 years. Using Heikezijing, a japonica rice landrace with broad resistance against rice blast and Suyunuo recombinant inbred lines, 13 QTLs were identified to be effective against only one M. oryzae isolate, while the other seven QTLs may be non-isolate-specific because each QTL confers resistance to more than one isolate [89]. By evaluating the disease reactions of 60 US weedy rice accessions with 14 M. oryzae isolates, 28 resistant QTLs were identified, of which three loci contribute to non-isolate-specific resistance [90]. With a combination of genome-wide association studies (GWAS) and interval mapping analyses, 51 QTLs against Xoo and Xoc were identified in multiparent advanced generation intercross populations, including 11 broad-spectrum resistance, three pathovar-specific, and 37 isolate-specific QTLs [88]. A GWAS analysis of 236 diverse rice accessions revealed 12 QTLs, of which two QTLs showed broad-spectrum resistance to Xoc [91]. Recently, 147 SNP associated with Xoo resistance were identified in 222 predominantly Thai rice accessions; however, the significantly associated SNP only occurred across a maximum of five Xoo isolates [92]. In rice, most resistance QTLs are conditioned to different populations and different QTL mapping analyses, which makes it difficult to handpick suitable QTL candidates for breeding programs with multiple resistances. To integrate QTL from different studies, a meta-analysis of QTLs represents a good approach. Using meta-analysis, 48 meta-QTLs were obtained from 27 studies, of which MQTL8.1 and MGTL2.5 were associated with resistance to rice blast, sheath blight, and bacterial leaf blight [93]. Resistant QTLs confer a partial but frequently referred resistance to broad-spectrum pathogen isolates or diverse pathogens, which are considered as effective resources for breeding to achieve broad-spectrum resistance [1]. However, the results from all these studies clearly showed that most QTLs confer isolate-and/or pathogen-specific resistance; in other words, not all resistance QTLs are broad-spectrum.
Moreover, there are few broad-spectrum resistant QTLs available for crop improvement programs, which results in the breeder needing a longer time and higher cost to pyramid QTLs to obtain broad-spectrum disease-resistant varieties.
All aforementioned studies give excellent examples of pyramiding R genes to obtain broad-spectrum resistance rice. However, it is still very difficult to obtain broad-spectrum disease resistant varieties by polymerization breeding for several reasons [111,112]. The first is that only a few R genes have been successfully used for molecular breeding processes. The existence of R genes containing resistant germplasms with excellent comprehensive traits is an important perquisite for breeding application. Secondly, the utilization of R genes in main modern rice varieties and the effectiveness of R genes in different ricecultivating regions are still not very clear. Thirdly, the resistance effect of pyramiding different R genes may not be a simple accumulation of resistance spectrum and improvement in the resistance level; meanwhile, with the increase in the number of pyramided R genes via traditional genetics approaches, the workload of breeding, time consumption, and linkage drag with unacceptable traits increase. Therefore, an accurate understanding of R gene utilization and the establishment of high-throughput molecular breeding methods to create R genes harboring resistant germplasms without linkage drag are important steps to overcome these difficult points, so as to improve broad-spectrum resistance breeding in the future.

Engineering Broad-Spectrum Disease Resistance Rice by Editing Susceptibility and Executor R Genes Is a New Method with Broad Application Prospects
During the plant-pathogen interaction, phytopathogens evolve to exploit the susceptibility (S) genes of plant to facilitate their infection. These S genes are associated with host recognition, penetration, pathogen proliferation and spread, or negative regulation of immune signals [113]. Disrupting these S genes can lead to enhanced resistance or reduced compatibility and, consequently, expand resources for broad-spectrum disease resistance. To date, many S genes have been identified in rice, such as Pi21, Xa5, Xa13/OsSWEET11, Xa25/OsSWEET13, and Xa41/OsSWEET14 [29,[114][115][116][117][118][119][120][121]. Recent advances in genome editing technologies, such as the CRISPR (clustered regularly interspaced short palindromic repeats)/Cas9 (CRISPR-associated protein 9)-mediated gene editing system, have greatly accelerated the generation of new resistant rice through genetic manipulation of S genes [113,122].
The most reported examples of editing S genes in rice can be found in research related to Xoo resistance. During the infection process of Xoo, abundant transcription activator-like effectors (TALEs), which are the major virulence factors and compatibility determinants, are secreted into rice cells. Most TALEs bind to the cis-element effector-binding elements (EBEs) in the promoter of S gene and reprogram their transcription to promote disease.
For instance, the TALEs PthXo1, PthXo2, and PthXo3/AvrXa7/TalC/TalF bind the EBEs in the promoters of rice OsSWEET11, OsSWEET13, and OsSWEET14 genes, respectively (Figure 2a) [122]. Editing the EBEs of S genes OsSWEET11, OsSWEET13, and OsSWEET14 in rice varieties japonica Kitaake and indica IR64 and Ciherang-sub1 resulted in loss of induction of these S genes by Xoo and broad-spectrum resistance against Xoo [123][124][125][126]. Similar strategies were used in the modification of S genes, Pi21, Bsr-d1, and Xa5 to obtain broad-spectrum resistance rice against Xoo and M. oryzae [127]. In addition, it was well summarized that editing the open reading frame of susceptibility defense regulators could obtain broad-spectrum resistance rice plants in the Wang's review (Figure 2b) [128].
Contrary to the interaction between TALEs and rice S genes, the usually suppressed executor R genes, such as Xa10 and Xa23, are transcriptionally activated by TALEs to restrict the growth of Xoo [28]. Using an in-depth understanding of the mechanism underlying the interaction between TALEs and executor R genes, a new strategy for engineering broadspectrum bacterial blight resistance through CRISPR/Cas9-mediated precise homology directed repair was proposed. Using this strategy, the EBE AvrXa23 , which is bound by TALE AvrXa23 to activate the expression of Xa23, was inserted into the promoter region of the susceptible xa23 allele in the susceptible rice cultivar, resulting in a resistant variety (Figure 2c) [129]. This is a significant expansion to the application of executor R genes and a new genome editing strategy in improving rice disease resistance.
EBEs in the promoters of rice OsSWEET11, OsSWEET13, and OsSWEET14 genes, respectively (Figure 2a) [122]. Editing the EBEs of S genes OsSWEET11, OsSWEET13, and OsSWEET14 in rice varieties japonica Kitaake and indica IR64 and Ciherang-sub1 resulted in loss of induction of these S genes by Xoo and broad-spectrum resistance against Xoo [123][124][125][126]. Similar strategies were used in the modification of S genes, Pi21, Bsr-d1, and Xa5 to obtain broad-spectrum resistance rice against Xoo and M. oryzae [127]. In addition, it was well summarized that editing the open reading frame of susceptibility defense regulators could obtain broad-spectrum resistance rice plants in the Wang's review (Figure 2b) [128].
Contrary to the interaction between TALEs and rice S genes, the usually suppressed executor R genes, such as Xa10 and Xa23, are transcriptionally activated by TALEs to restrict the growth of Xoo [28]. Using an in-depth understanding of the mechanism underlying the interaction between TALEs and executor R genes, a new strategy for engineering broad-spectrum bacterial blight resistance through CRISPR/Cas9-mediated precise homology directed repair was proposed. Using this strategy, the EBEAvrXa23, which is bound by TALE AvrXa23 to activate the expression of Xa23, was inserted into the promoter region of the susceptible xa23 allele in the susceptible rice cultivar, resulting in a resistant variety (Figure 2c) [129]. This is a significant expansion to the application of executor R genes and a new genome editing strategy in improving rice disease resistance.

Transgenic Rice Expressing Genes from Other Species Shows Broad-Spectrum Disease Resistance
Along with pyramiding resistance genes and editing susceptibility genes in rice, development of transgenic rice plants by expressing defense genes from other species is an appropriate approach to control pathogens, especial in the absence of a resistant germplasm. For instance, the Arabidopsis NPR1 protein (non-expressor of PR1) is a key regulator in the signal transduction pathway leading to the activation of SAR, which is a broad-spectrum resistance response upon exposure to a pathogen [130]. Constitutive expression of the AtNPR1 gene in rice leads to high resistance but growth and agronomic

Transgenic Rice Expressing Genes from Other Species Shows Broad-Spectrum Disease Resistance
Along with pyramiding resistance genes and editing susceptibility genes in rice, development of transgenic rice plants by expressing defense genes from other species is an appropriate approach to control pathogens, especial in the absence of a resistant germplasm. For instance, the Arabidopsis NPR1 protein (non-expressor of PR1) is a key regulator in the signal transduction pathway leading to the activation of SAR, which is a broad-spectrum resistance response upon exposure to a pathogen [130]. Constitutive expression of the AtNPR1 gene in rice leads to high resistance but growth and agronomic trait defects.
To overcome this problem, different strategies were developed by two groups [130,131]. Eventually, broad-spectrum resistant rice plants without a fitness cost were obtained by expression of AtNPR1 driven by green tissue-specific promoter or pathogen-responsive upstream open reading frames of key immune regulators TBF1 [130,131]. Similarly, transgenic rice lines expressing the auto-activated NLR genes RPS2 and RPM1 (D505V) from Arabidopsis conferred broad-spectrum resistance to pathogens M. oryzae and Xoo via early and strong induction of ROS, callose deposition, and expression of defense-related genes. These RPS2 and RPM1 cases revealed that auto-activated NLRs are a promising resource for breeding crops with broad-spectrum resistance, and they provide new insights for engineering disease resistance [132]. In addition to genes from Arabidopsis, transgenic rice plants expressing resistant Lr34 allele from wheat showed increased resistance against multiple isolates of the hemibiotrophic pathogen M. oryzae by delaying invasive hyphal growth [133]. In another example, transgenic rice lines expressing the isoflavone synthase (GmIFS1) gene from soybean contributed to the synthesis of isoflavone (genistein) to promote M. oryzae resistance, indicating that the synthesis of heterologous secondary metabolites, such as isoflavone, is a good way to develop blast resistance in rice [134]. As such, we believe that engineering resistant rice through ectopic transcription of defense genes cloud be a broadly applicable new strategy, which may lead to reduced use of pesticides and lightening the selection pressure of resistance pathogens.

Conclusions
In the past 10 years, several broad-spectrum R genes, defense regulators, and QTLs were identified in rice with resistance against two or more types of pathogen species or many isolates of the same pathogen species. This emerging knowledge of broadspectrum resistance genes formulates efficient ways to best use these genetic resources for crop improvement via biotechnological approaches. However, there are still many gaps in our knowledge of the mechanisms underlying broad-spectrum resistance. To reveal these mechanisms, more research about the interactions between the host R protein and pathogen effectors, as well as the cooperation among these broad-spectrum resistance genes, is required in future. Furthermore, there is still a long way to using these genes to create broad-spectrum disease-resistant varieties. An accurate understanding of R gene utilization, the establishment of high-throughput molecular breeding methods to create R genes harboring resistant germplasms without linkage drag, and an investigation of new strategies for using defense regulator genes without a yield penalty will be helpful for improving broad-spectrum resistance breeding in the future.