Disease Resistance and Susceptibility Genes to Bacterial Blight of Rice

Rice ( Oryza sativa L.) is a valuable resource for understanding the complex processes controlling yield and value-added traits. Bacterial blight (BB) is a vascular disease of rice, caused by strains of Xanthomonas oryzae pv. oryzae ( Xoo ) and provides insight, both practical and basic, into the concepts of susceptibility and resistance. Basic knowledge has been empirically and, more recently, intentionally exploited for broad and durable resistance to the disease. Bacterial blight involves representatives of most classes of resistance genes ( R genes) and pathways for basal plant immunity. The study of BB also revealed novelties not observed in other models, possibly due to the long history of rice cultivation and the constant disease pressure. Conspicuous are the recessive R genes that target the notorious type III Transcription Activator-like effectors (TALes) of Xoo . Results indicate that pathogen and host are currently in a battle over a small patch of ground involving TALes function. At the same time, analyses of rice disease physiology are adding to a growing body of knowledge for plant disease processes and to how these processes are intertwined with disease susceptibility. The basic processes of BB present rich targets for the rapid advances in genome editing.


Introduction
World population is expected to rise beyond 9 billion by 2050 [1]. Rice (Oryza sativa) is a staple food crop world-wide, providing about one fifth of the calories consumed by humans [2]. In particular, rice accounts for 35-75% of the calories consumed by more than 3 billion in Asian countries alone and planted on approximately 154 million hectares land annually [3]. Crop protection and food security go hand in hand, and breeding for resistance against crop diseases remains the essential ingredient for food security. Due to the labor-intensive nature of breeding, integrated disease control is often reduced to mere chemical control, leaving the very purpose of this environment-friendly approach in limbo. Advances in molecular tools in crop breeding, however, makes breeding an increasing sustainable effort in staying ahead of pathogen adaptation [4]. Bacterial blight (BB) of rice is a widespread vascular disease caused by Xanthomonas oryzae pv. oryzae (Xoo). Epidemics can severely reduce grain yield due to collapse of the entire crop [5]. BB was first characterized in the late nineteenth century [6]. Introduction of resistance (R) genes into rice cultivars is considered as the best option for Xoo management. A total of 42 R genes have been identified in rice against Xoo, and the number continues to grow [7][8][9]. Due to co-evolution and selection pressure between Xoo and rice, these R genes are selective in their efficiency against specific Xoo strains or races, which are sets of strains that share incompatibility on defined sets of R genes [10].

Post genomic era and rice grain protection
Advancements in genomics, referring here to DNA and RNA analyses, is as beneficial to crop protection as is to other discipline of biology. Rice MetaSysB, an open source which provides detailed information about BB-responsive genes, is based on the global expression analysis. The database provided 7475 unique genes and 5375 simple sequence repeats, which were responsive to Xoo in rice [11]. Such information is based on the compatible and incompatible rice-Xoo interactions. In another example, 454 and 498 differentially expressed genes were reported as exemplified by the incompatible and compatible rice-Xoo interactions, respectively, using cDNA microarray [12]. Genomics also provides functional information of genes up-and downstream of candidate resistance genes in the defense signal pathway, as is done in near-isogenic rice lines introgressed with Xa39, an as yet uncharacterized BB resistance gene [13].
Multiple rice and Xoo genomes have been sequenced, either in draft or complete form [14][15][16][17][18][19][20][21][22][23], paving the way to identify functional connections between host and pathogen genes. The functional validation of the candidate genes is helping develop new rice varieties by introduction of the gene of interest through traditional breeding, marker assisted breeding, or genetic engineering approaches [3]. BB disease resistance is overcome by the emergence of more virulent strains of Xoo. Whole genome sequencing of 100 Xoo strains from India revealed that these strains were distinct from African and US Xoo strains [24]. Based on the reaction towards ten major resistance genes of rice, 46 out of the 100 strains were grouped into 11 pathotypes [24].

The genetic context of rice-Xoo interaction
Many BB-resistance genes in modern rice germplasm were selected long before the concepts of modern plant breeding were established, and a rich assortment of major dominant and recessive R genes has been identified by genetic and molecular studies ( Table 1).
Perhaps the best known of these genes, Xa21 represents the receptor kinase (RLK) class of R genes. Xa21 was originally introgressed into rice from the related species O. longistaminata and confers resistance to a broad range of Xoo strains [25]. Xa26, another cloned member of RLK gene family, also confers broad resistance with a somewhat different strain profile [26]. The cognate elicitor for Xa21 has been reported [27]. However, for Xa26 has not been identified.
RLKs play a central role in disease immunity pathways in plants, largely via the characterization of the bacterial flagellin receptor FLS2 and the related receptor EFR in Arabidopsis [28,29]. A typical RLK consists of an extracellular receptor domain comprising of leucine-rich repeats (LRRs), a transmembrane domain, and an intracellular kinase domain [30]. As a class, RLKs have great potential for enhancing resistance to BB in rice and in other disease complexes of crop plants Xa21, Xa26, and other RLKs represent genetic components of the pathogenassociated molecular patterns (PAMPs)-triggered immunity (PTI) surveillance pathway in rice. Improvements in the rationale design of RLK receptor specificities, and screening for novel genes in germplasm or wild relatives could lead to general application for broad and durable resistance.
The nucleotide binding site-LRR (NBS-LRR) is another large class of R gene, represented in rice toward Xoo by Xa1 and Xo1 [31][32][33]. XA1 and XO1 recognize multiple TALe, and Xoo strains have adapted TALes, the so-called iTALes, that are truncated and inhibit the function of XA1 and XO1 [32,34].
Specific TALe-dependent R genes governing dominant resistance in rice against Xoo are known as executor (E) genes. E genes are distinct from classical R genes, whose transcriptional activation by TALes of Xoo trigger immunity, leading to dominant resistance [35]. Xa27 represents the E genes class of dominant R genes and confers broad resistance to BB in rice [36]. Although not expressed in susceptible host, Xa27 is expressed only upon inoculation with Xoo strains harboring the TALe gene avrXa27 [37]. The protein is localized to apoplastic space, cell membrane and cell wall, and when expressed under a pathogen-nonspecific inducible rice OsPR1 promoter, conferred constitutive resistance to both compatible and incompatible Protecting Rice Grains in the Post-Genomic Era strains alike [37]. The rice R genes Xa10 and Xa23 have similar requirements for the transcription activation domain and nuclear localization sequence (NLS) motifs of the corresponding TALes for their induction [38,39]. Xa4 is the latest and, again, an unusual R gene of rice to be characterized. The protein is a wall-associated kinase (WAK) and provides attributes other than enhanced resistance. Rice plants with XA4 are shorter and stiffer in comparison to plants lacking the gene [40]. Xa4 is race-specific, meaning many strains of Xoo are compatible on plants with Xa4. How Xa4 functions in resistance is unknown at present.

SWEET genes and recessive resistance
A class of major TALe-dependent susceptibility (S) genes for BB in rice encodes sugar transporters, thereby named as SWEET gene family [41]. Specific TALes, referred to as major TALes, transcriptionally activate the corresponding SWEET genes in rice during infection to promote the disease in a gene-for-gene susceptibility manner [42]. Although at least five SWEET genes of the clade III members can function as an S gene in BB, only three members are known to be targeted by extant strains of Xoo [42][43][44][45][46][47]. A member of the SWEET gene family, OsSWEET14, is induced by multiple distinct TALes, which include AvrXa7, PthXo3, Tal5 and TalC and are present in strains of different geographic origins and genetic lineages [43,45,46]. Similarly, PthXo2 drives OsSWEET13 expression in the susceptible rice variety IR24 [44], and OsSWEET11 is induced by the cognate PthXo1 [42]. The typical TALe possesses a central repetitive domain, a nuclear localization signal domain, and a transcription activation domain. The repetitive domain is responsible for binding of the TALe to a sequence motif called the effector binding element (EBE), which is commonly located in the promoter region of the respective S gene.
Mutated S gene alleles are proposed to be potentially more durable than dominant R genes [48,49]. Identifying the promoter variant alleles of major S genes has been proposed in breeding for BB resistance [42,47,[50][51][52][53]. Recessive resistance is due to the cognate TALe cannot bind to the promoter variants of the S gene. The gene xa13, for example, is a recessive resistance insertion allele of 14.8 kb DNA fragment in the promoter of OsSWEET11 [42,47]. OsSWEET11 encodes a protein related to MtN3 encoding nodulin 3 (N3) protein of Medicago truncatula. The gene was originally named Os8N3 due to its location on rice chromosome 8 and the similarity to MtN3 [42]. The critical difference between resistant (xa13/xa13) and susceptible plants is the elevated expression of OsSWEET11 during infection in otherwise susceptible plant genotypes [42]. RNAi-mediated silencing of OsSWEET11 plants was similarly resistant to Xoo strains that are solely dependent on PthXo1 for SWEET induction. Silenced plants, but not promoter variants, showed low pollen viability, corroborating the fact that Xoo hijacked otherwise developmentally important genes in rice for pathogenicity [42,47]. Similarly, the TALe PthXo2 cannot bind to the EBE of xa25, a recessive allele of OsSWEET13, or the EBE region of OsSWEET13 in japonica rice cultivars, owing to single nucleotide polymorphisms in the respective EBEs [51, 52].
The gene xa13 is a naturally occurring allele, actually a series of alleles that protects the plant from a genetic disease vulnerability in the plant developmental pathways [42,47]. However, xa13 is not a broad resistance provided in comparison to Xa21, Xa27 and xa5), and many strains from China, Philippines, Japan and Korea are compatible on xa13 lines [51]. Compatibility is derived by acquisition of major TALes that target alternative SWEET promoters [43]. As yet, not major TALe has been identified that replaces PthXo1 for OsSWEET11 expression. The gene xa5 also affects TALe-dependent function but does not act at a specific SWEET gene. The recessive allele encodes a variant of the γ or small subunit of the transcription factor TFIIA [54, 55], which confers broad resistance. The gene differs from the susceptible allele by a single codon substitution of valine at position 39 to glutamic acid. TFIIA, consists of α, β, and γ subunits, and is involved in stabilizing the binding of the TATA box binding protein complex (TFIID) to the TATA box of gene promoters. The TFIIA components are highly conserved across the eukaryotes. Rice has two loci for TFIIAγ-one gene is on chromosome 5 (TFIIAγ5, xa5)

Implication of interactions between TALes and the corresponding host genes
Due to the large reservoir of TALes in each strain of Xoo and the diverse roles of TALes in pathogenesis, the BB of rice represents an excellent plant/pathogen system for studying the biology of TALes. The apparent reason for the broad activity of Xa27 and Xa23 is the presence of the cognate TALes avrXa27 and avrXa23 in a large number of strains from southeast Asia, including Korea, China, Japan and the Philippines [37,39]. On the other hand, the loss of avrXa27, avrXa23, or avrXa10, for that matter, does not appear to have an apparent fitness cost to the pathogen, and populations of Xoo may lose avrXa27 if Xa27 is widely deployed [37][38][39]. AvrXa7 is an important virulence factor for some strains of Xoo, and strains with AvrXa7 are incompatible on rice lines harboring the Xa7. In this case, loss of avrXa7, which is a major TALe for OsSWEET14, may result in strains that are weakly virulent or, essentially, nonpathogenic, if no other SWEET inducing TALes are present [43,58]. A variety of other TALe genes are present in Xoo populations that can restore full virulence to strains missing avrXa7 [59]. Evasion of Xa7-mediated resistance is possible by loss of the gene, rearrangement of the central repeats or recombination among different TALe genes [60,61]. However, despite rapid adaptation of bacteria by genetic changes and gene flow, field studies in the Philippines indicated that deployment of Xa7 was durable in test plots for more than 10 years [62]. Therefore, strains may have other limitations due to geographical location or rice genotype. Nevertheless, pyramiding broadly effective R genes with cognate TALes that are wide-spread in the pathogen populations should provide a degree of broad and durable resistance.
In the case of xa13, induction of the dominant allele SWEET11 is mediated by the TALe PthXo1 [42]. However, strains of Xoo that solely rely on PthXo1 cannot induce xa13 allele, and rice homozygous for xa13 is symptomless. xa13-dependent recessive resistance is phenotypically and qualitatively different from resistance provided by the dominant R gene Xa7 [42,63]. Quantitatively, however, resistance mediated by xa13 and Xa7 are approximately equal with respect to bacterial growth and lesion length [42,58,64]. Xa7 resistance is the result of the presence of the appropriate AvrXa7 in the pathogen and dominant, while xa13 resistance is dependent on the absence of an effective virulence factor and recessive. The mechanism of XA7 mediated resistance is as yet unknown.
Type III effectors, in general, are hypothesized to interfere with host defense and defense signaling mechanisms. Strains of Xoo have other type III effectors, differing from TALes, and, therefore, not entirely dependent on TALes for suppression of host defenses [65]. Xoo strains lacking major TALes are still capable of causing water-soaking, if syringe inoculated, which is in contrast to type III secretion system (Hypersensitive reaction/pathogenicity or Hrp − ) mutants. Hrp − mutant strains are incapable of secreting any type III effectors, including TALes, and are virtually symptomless [66]. The mechanism by which SWEET transporters condition susceptibility is unknown. One hypothesis is that the transporters allow cells to leak sucrose, providing the pathogen with nutrients. SWEET function may interfere with normal plant defense functions or, possibly, allow transport of other nutrients or disease promoting compounds [41]. However, little empirical evidence for the nutrition model exists at present.
Sequencing of Xoo genomes has revealed the full complement of TALes is now known [17][18][19][20][21][22][23]. The individual TALe genes are distinguishable on the basis of the number of repeats in the central repetitive region and by polymorphisms within each repeat sequence, particularly, at the 12 th and 13 th codons. Strains of the Asian lineage contain upwards of 16-19 TALe genes in each genome [18]. The large numbers of TALe genes in these species may reflect the evolutionary investment in utilizing the TALes for virulence and are essential, to the ecological niche these bacteria occupy. The maintenance of a large repertoire of TALe genes may increase the frequency of recombination between, and diversity of TALecgenes within the pathogen population [60]. Pathogen may then adapt faster to the changing host genotypes as exemplified by the appearance of pthXo5, which avoids Xa7 recognition and appears to be a hybrid between avrXa7 and pthXo6 [61].
Not all TALE genes of Xoo, however, are just substrates for new major TALEs. Two other TALE genes from PXO99 strain of Xoo, in addition to pthXo1, contribute to virulence, known to elevate the expression of two host genes distinct from SWEET11. PthXo6 elevates the expression of OsTFX1, which contributes to approximately 35% of the disease [67]. Many strains induce OsTFX1. The gene pthXo7 of PXO99 elevates the expression of OsTFIIAγ1 and would appear to be an adaption to host genotypes containing the xa5 allele of TFIIAγ5 [67]. However, introduction of pthXo7 to other strains does not restore full virulence on xa5/xa5 plants and may provide only an incremental fitness benefit [67]. All Asian strains also carry a set of truncated TALes, the inhibitory or iTALes, which function to suppress Xa1mediated resistance [32].

Executor R genes and super promoters
Xa10, Xa23 and Xa27 are representatives of the new class of E genes, so-named because the induction of these genes executes a response of programed cell death (PCD) in the host. Xa10 induced PCD in plant species rice and N. benthamiana, and mammalian HeLa cells [38]. No cognate S genes for AvrXa10, AvrXa23, or AvrXa27 in compatible host cultivars have been reported, though the presence of AvrXa27 and AvrXa23 in many extant strains of Xoo may portend either a defeated function or an unknown cryptic function in S gene expression. Nonetheless, E genes hold great potential for broad and durable resistance in rice against extant Xoo population. A super promoter consisting of multiple EBEs, corresponding to specific TALes in extant population of Xoo, have been constructed (Figure 1). [68][69][70]. Addition of multiple EBEs to a pathogen strain specific rice BB resistance gene makes it effective against additional strains of Xoo. The EBEs of TALes PthXo1, PthXo6 and Tal9a when conjugated to E gene Xa27, showed resistance against PXO99 and a derivative strain lacking AvrXa27 [68]. A similar scenario was accomplished using E gene Xa10 [69]. The study suggested that broad-spectrum and potentially durable resistance is possible by stable integration of an E gene engineered in a way to respond to multiple TALes from different strains or even different pathogens. Design of a super promoter, however, needs to be done carefully. Risk that an added EBE might coincidently contain a cis regulatory element could induce the E gene expression in response to particular stimuli and cause cell death without challenge by TALes. Amended promoters should be tested thoroughly before deployment.

Targeted genome regulation and editing
Central to TALe function is the discovery of the DNA recognition cipher of TALEs [71,72]. The central domain of a TALe, also known as binding domain, consists of variable number of tandem repeats, each consisting 33-35 amino acid residues. The 12 th and 13 th amino acid residues (known as repeat variable di-residues, RVDs) of each repeat preferentially binds to the respective nucleotides in the EBEs of target gene, such that HD, NG, NI and NN bind to C, T, A, and G, respectively in the effector binding elements (EBEs) of the promoter of a target gene [71][72][73]. The TALe recognition code allowed custom-engineer of DNA binding domains, also called designer TALes (dTALes), with novel specificity to the user-chosen DNA sequences [74][75][76]. dTALes provide a useful tool box to transiently activate host genes of interest for their functional analysis and assess the associated effect on host phenotype and physiology during rice-Xoo interaction. TALENs are fusions between dTALes and the nuclease domain of restriction enzyme FokI [77][78][79][80]. Other C-terminal domains have also been used [81]. Target site recognition and TALEN dimerization triggers a double-strand break (DSB) and generates small random insertions or deletions at the cleavage site, resulting in an edited sequence. CRISPR-Cas editing approaches have circumvented the need to construct dTALes and achieved wide general use, including editing of rice genes [82][83][84].

Prospects for engineered broad and durable resistance in rice to BB
Traditional resistance breeding has identified many useful R genes and introgressed the genes into elite cultivars. Further, development of molecular markers allows the pyramiding of multiple genes into single lines. The development of designer TALENs and CRISPR-Cas genome editing brings greater flexibility and rapidity to the development of resistant germplasm. A continuous provision of novel R genes in breeding programs is possible. Of course, the adoption and utility of different approaches is dependent on the regulatory climate. Introduction of novel or alien genes may be prohibitive in the foreseeable future. Classification of genome editing techniques will also vary depending on the individual country. In the rice system, our understanding allows numerous approaches for the enhancement of resistance beyond classical breeding. TALe biology, specifically, can be exploited (Figure 2). Least intrusive is targeted genome editing of S genes. OsSWEET14 is targeted by unrelated TALEs, AvrXa7, PthXo3, Tal5 and TalC from different Xoo strains and which in some cases overlap their EBEs [43,45,46]. OsSWEET14 was made unresponsive to TALEs AvrXa7 and Tal5, when their respective EBEs were mutated using TALENs in otherwise susceptible rice cv. Kitake [85,86]. Thus, recessive resistance obtained by the genome editing of OsSWEET14 is expected to be broad and contribute to durability given the apparent few major TALes in the extant population. Future efforts will be to target all EBE/S gene combinations in single elite lines. Fusion of EBEs to a variety of R and E genes has  been demonstrated to provide resistance [68,87]. The functional specificity of an E gene can be broadened by linkage to general inducible defense genes [69,88]. Approaches are not limited to TALe-associated responses. The RLK immunity receptor EFR from Arabidopsis [89,90], as well as XA21/EFR fusion proteins function in rice [91]. Thus, the sky is the limit for the engineering of broad and durable resistance in rice to BB.