Genomic and Molecular Perspectives of Host-pathogen Interaction and Resistance Strategies against White Rust in Oilseed Mustard

Abstract

Oilseed brassicas stand as the second most valuable source of vegetable oil and the third most traded one across the globe. However, the yield can be severely affected by infections caused by phytopathogens. White rust is a major oomycete disease of oilseed brassicas resulting in up to 60% yield loss globally. So far, success in the development of oomycete resistant Brassicas through conventional breeding has been limited. Hence, there is an imperative need to blend conventional and frontier biotechnological means to breed for improved crop protection and yield.

This review provides a deep insight into the white rust disease and explains the oomycete-plant molecular events with special reference to Albugo candida describing the role of effector molecules, A. candida secretome, and disease response mechanism along with nucleotide-binding leucine-rich repeat receptor (NLR) signaling. Based on these facts, we further discussed the recent progress and future scopes of genomic approaches to transfer white rust resistance in the susceptible varieties of oilseed brassicas, while elucidating the role of resistance and susceptibility genes. Novel genomic technologies have been widely used in crop sustainability by deploying resistance in the host. Enrichment of NLR repertoire, over-expression of R genes, silencing of avirulent and disease susceptibility genes through RNA interference and CRSPR-Cas are technologies which have been successfully applied against pathogen-resistance mechanism. The article provides new insight into Albugo and Brassica genomics which could be useful for producing high yielding and WR resistant oilseed cultivars across the globe.

1. introduction

Oilseed crops have been the backbone of several agricultural economies from antiquity and play a prominent role in agricultural industries and trade throughout the world [1]. Indian mustard (Brassica juncea. (L.) Czern.), a member of the Brassicaceae family, is an economically important oilseed crop cultivated globally. It is a major oilseed crop of Indian subcontinent and the second most important source of edible oil, contributing about 30% to the total edible oilseeds production in the country. It occupies 80% of the total area of oilseed Brassica cultivation in India [2]. The yield is limited due to environmental adaptability challenges and damage caused by phytopathogens. One of the major diseases of crucifer includes white rust caused by Albugo candida (Pers.) Kuntze resulting in up to 60% yield loss of Indian mustard globally depending upon host genotype, planting time, plant population, nutrition and climatic conditions across the globe [3].

Disease resistance to oomycete pathogens has been a major target of plant disease management programs, and also a prime focus of genomics research to identify the major resistance genes and their molecular basis of resistance.

2. WHITE RUST: DISEASE BIOLOGY

White Rust (WR), white blister and white blister rust are the common names of the disease caused by Albugo spp. in more than 400 species of plants worldwide [4-6]. Name of the disease is derived from the appearance of white pustules, due to enzymatic digestion of epidermal cell wall, on the surface of leaves and other aerial parts of the host [7]. The white pustules are mass of dehydrated sporangiospores that upon re-hydration in water droplets lead to stomatal infection [8]. It is a member of the eukaryotic oomycete order Albuginales under class oomycota, which constitutes exclusively obligate biotrophic parasites with a broad plant host range [9-13]. A. candida exhibits obligate biotrophic nutrition being completely reliant on host tissues. A. candida reproduces by asexual sporangia or zoospores and extremely resistant thick-walled sexual oospores (Fig. 1). In all species of WR pathogens, oospores are the primary source of inoculum [3]. The oospores are responsible for long-term survival in plant debris and liberated when the host tissues decay [10-13]. The presence of oospores in crop residues and/or perennial mycelium in living host (including weed) tissue enables the pathogen to survive between host growing seasons [14]. Moisture on host surfaces is essential for the germination of sporangia and infection by the zoospores. The most likely primary infection sites are the emerging cotyledons of host plants [3]. Albugo sp. enter via stomata, form intercellular hyphae, penetrate the plant cell wall and invaginate the plant plasma membrane with haustoria in order to take up plant nutrients and release effector proteins into host cells [12]. When zoospores come in contact with the plant leaf surface, they settle in stomata, encyst and produce a germ tube, which extends into the sub-stomatal chamber and penetrates the host cell. A primary vesicle forms in the host cell, which enables further development of intercellular hyphae in a susceptible interaction [8]. When an Albugo infection matures, zoosporangia forcibly rupture the plant epidermis and further enzymatic digestion results in characteristic “white blister” pustules [14]. Both local and systemic types of expression characterize the disease. Local infection appears as white or creamy yellow pustules or “blisters” on leaves and stems. Systemic infection results in abnormal growth, distortion of inflorescence and sterility of flowers commonly called staghead, which appears as a result of hypertrophy and hyperplasia. Apart from A. candida, which infects oil yielding Brassicas and crucifer vegetables, several other Albugo species are also well-known plant pathogens causing huge yield loss in field crops of economic importance, e.g., Albugo tragopogonis in sunflower, Albugo ipomoeaepandurate in sweet potato, and Albugo occidentalis in spinach. A. candida is an obligate biotrophic homothallic oomycete pathogen of white rust [15]. As per the molecular investigations, the genus Albugo constitutes ∼50 (usually) specialist pathogens such as A. laibachii in Arabidopsis thaliana, and A. candida in B. juncea [16-19]. The impact of the disease is very high in Indian sub-continent as almost all released lines grown commercially in India are susceptible to this disease.

Fig. (1).

A. candida life cycle representing the sexual and asexual cycle ([3]). (A higher resolution / colour version of this figure is available in the electronic copy of the article).

A. candida forms 24 physiological races as reported to date which infect over 200 species of plants in 63 genera from the families of Brassicaceae, Cleomaceae and Capparaceae, each of which specializes in different host species, out of which at least 10 specialize in different Brassicaceae species [20-30]. Among the identified races, race 2 (Ac2VRR) (the ‘RR’ suffix was added to the pathotype name Ac2v to indicate the standard isolate chosen from the collection of the late Dr. Roger Rimmer) causes a severe annual loss of oilseed mustard (Brassica juncea [L.] Czern. and Coss.) in India, Canada and Australia and also infects some genotypes of other Brassica spp., including oilseed turnip rape (Brassica rapa L.) [21, 30-32]. Race 1 (Ac1) infects Raphanus sativus, Race 4 (Ac4) infects Capsella bursa-pastoris, Race 5 (Ac5) infects Sisymbrium officinale, and Race 6 (Ac6) infects Rorippa islandica. Race 7 (Ac7) is largely restricted to B. rapa, but has also been reported to cause disease in some cultivars of B. napus [33], and some genotypes of B. juncea [31]. The Race 9 (Ac9) infects B. oleracea.

3. EFFECTOR MOLECULES – MAJOR WEAPON TO ESTABLISH INFECTION

The genomes of several oomycetes, including Phytophthora infestans, Phytophthora ramorum, Phytophthora sojae, Pythium ultimum, Albugo laibachii, A. candida, and Hyaloperonospora arabidopsidis, have been sequenced and found to encode hundreds of effector candidate proteins [12]. They are an interesting clad of microorganisms to be studied with 18 Mb to 240 Mb genome and rich secretome [8]. A. candida with a 45.3 Mb genome is significantly smaller than a comparable significant biotroph such as H. arabidopsidis. In the early phases of the interaction, the invading oomycete needs to deal with biochemical barriers in the plant apoplast to overcome the plant immunity. Both pathogen and host secrete proteins and metabolites to control the extracellular environment.

An oomycete-plant interaction is characterized by molecular coevolution, with each side battling for control over the other. Being obligate parasites they are highly specialized in suppressing the Pattern Triggered Immunity (PTI) in plants and colonizing the host keeping it alive. Albugo sp. enter into the plant cell via stomata, invaginating the plasma membrane with haustoria to take up the plant nutrients and release the effector proteins into the host cells [13]. Effector protein further leads to infection by suppressing host plant defense responses as well as adapting the host metabolism [33]. Effectors and elicitors delivered to the apoplast include protease inhibitors, glucanase inhibitors, small cysteine‐rich proteins, necrosis‐like proteins (NEPs) and elicitins [34, 35].

“Effectors” are secreted protein molecules that help both the invasion and the propagation of the pathogen by suppressing host plant defense responses as well as adapting the host metabolism [36]. They are studied to alter host cell structure and function, thereby facilitating infection (virulence factors or toxins) and/or triggering defense responses (avirulence factors or elicitors). Dawkins (1999), referred to the effectors as “parasite genes having phenotypic expression in host bodies and behavior” [37]. Pathogen utilizes the effector proteins to interfere with host cellular mechanisms by suppressing the host immunity for successful infection. [38-41]. The secretion of effectors is the prime mode of infection shown by the obligate parasites. Effectors can be classified into apoplastic or extracellular and cytoplasmic or intracellular on the basis of their target sites in the host plant [40, 42]. Apoplastic effectors are secreted into the plant extracellular space (host-pathogen interface), where they interact with extracellular targets and surface receptors. Three types of apoplastic effectors from oomycetes have been studied to interfere with plant processes; inhibitors of host enzymes, RGD (Arginine–Glycine–Aspartic acid) containing proteins disrupt cell wall-plasma membrane adhesions and toxins that lead to host cell death [43]. The suppression of plant defense, a consequence of PTI gets activated upon the invasion of pathogens, is key to successful infection of the host. It is evident that oomycete pathogens actively suppress innate immunity, as witnessed in A. thaliana by A. candida [44, 45]. Cytoplasmic effectors carry conserved peptide motifs at their terminals which aid them in their functioning and translocation. On the basis of the conserved motif, two important groups of translocated intracellular effectors, namely the ‘RxLR-effectors’ and the ‘Crinklers’ have been identified to be present abundantly in many plant pathogenic oomycetes. RxLR effectors (with conserved N-terminal Arg-Xaa-Leu-Arg motif) are the largest group of oomycete effector proteins [46]. Since 2005, more genome sequences of several oomycete pathogens have become available facilitating the computational analysis of the secretome in different oomycetes.

3.1. A. Candida Secretome

The de novo sequence assembly of A. candida revealed a small genome relative to other biotrophic oomycetes [8]. The draft genome sequence generated from the Ac2VRR, which infects the Indian mustard, reports the presence of minimal gene repertoire. Pen sequencing has helped us to assess the genome organization of A. candida and analyze the degree to which orthologous genes appear to have been retained relative to other oomycetes. Using BLAST, 5,975 putative orthologs to P. ultimum, 5,858 to P. infestans, 5,581 to P. sojae, 5,592 to P. ramorum, and 4,922 to H. arabidopsidis have been identified in A. candida. The genome analysis of the oomycete H. arabidopsidis reports a reduction in the number of genes encoding secretory pathogenicity proteins, enzymes for assimilation of inorganic nitrogen and sulphur representing a genomic signature for the evolution of obligate biotrophy [47]. Similarly, A. candida also lacks the assimilation enzymes such as nitrate, nitrite, and sulphite reductases, as a consequence, there is an absence of certain metabolic pathways [8]. These pathways are under investigation and not reported to date.

The Ac2VRR secretome appears to be of about 2/3 the size of the P. infestans secretome, a notable pathogenic oomycete, a responsible agent for Irish potato famine [48]. In contrast to the extensive family expansions observed in Phytophthora genomes, the gene families of the A. candida secretome have relatively few members. Sequencing of A. candida transcripts from infected host tissue and zoosporangia combined with genome-wide annotation revealed 15,824 predicted genes [8, 49]. Most of the predicted genes lack significant similarity with sequences from other oomycetes. There are twelve A. candida predicted genes that bore a similarity to other oomycetes, suggesting a shared evolutionary lineage to some of these genes amongst oomycetes [8]. A total of 929 proteins have been identified from the annotated set of A. candida proteins based on the criteria such as the presence of amino-terminal signal peptide, presence of a motif indicating the protein’s secretory function and the absence of additional transmembrane domain [8, 50]. The A. candida secretome consists of proteins without transmembrane domains, reflecting its wide host range. It secretes both apoplastic and cytoplasmic types of effectors. As already discussed, A. candida has a much smaller repertoire of pathogenicity-related proteins as compared to Phytophthora spp. and H. arabidopsidis including genes that encode RXLR effector proteins, CRINKLER-like genes, and elicitins. For example, the large A. candida secreted protein family consists of six Crinkler-like (CRN) proteins, whereas the largest family in Phytophthora and pathogen H. arabidopsidis has hundreds of members [47]. Some of the gene families of A. candida encoded proteins, resembling those associated with the elicitation of innate defense in plants and/or compatibility in plant-microbe interactions, and the function of several proteins are yet to be discovered.

In general, the A. candida genome harbors fewer genes in each of the major classes of PAMP or effector-like proteins. Cellulose-binding elicitor lectin (CBEL) are elicitors, and those of oomycetes are reported to play a role in adhesion to the plant cell surface. As reported by Links et al. (2011) [8], the genome of A. candida also contains two putative CBEL genes: Ac2VRR-CBEL1 and Ac2VRR-CBEL2. Ac2VRR-CBEL1 has a signal peptide, which may be responsible for its secretion primarily during the infection process. It also describes that Ac2VRR-CBEL2 does not contain a signal peptide indicating that it would not be secreted. Canonical structure of CBELs from Phytophthora species is a sec-dependent signal followed additionally by interleaved Cellulose Binding (CBD) and Apple Domains which mediate protein-protein or protein-carbohydrate interactions. Variation is observed amongst CBEL orthologs, which appears to be conserved within Oomycete genera. NEPs, capable of triggering plant cell death are not found in A. candida, whereas common among diverse groups of plant pathogens [51, 52].

A small group of genes in A. candida genome encodes secreted proteins with a variant RXLR motif (Ac-RXL). The translocation mechanism of these proteins in the plant is not fully clarified, but probably RxLR uptake may involve binding to a receptor on lipid rafts as shown for a host-targeted protein from the animal pathogenic oomycete Saprolegnia parasitica [53]. RxLRs and CRNs are known to defeat plant immune responses through many routes, which include reprogramming of host gene expression, altering RNA metabolism, and binding to host proteins involved in signaling [54]. Twenty-six predicted gene models with a putative sec-dependent signal peptide, non - homologous to known proteins, and had an Ac-RXL motif were estimated by string searches within the predicted proteins with amino-terminal signal peptides. Five candidate Ac-RXLs were reported to transiently induce necrosis when infiltrated into N. benthamiana [8]. CRNs have been proposed to be an ancient group of host-targeting proteins that evolved in the oomycete lineage, six gene models have been identified as putative CRNs. The A. candida CRNs form the largest secreted gene family in A. candida. CRN type effectors are not highly abundant in A. candida but are similar in number to P. ramorum. None of the domains present in the C- terminal as present in other oomycetes [51, 55, 56] were detected in Albugo, which supports the independent origin of Albugo effectors in the evolution of biotrophy. Systematically searched for novel host-targeting motifs amongst the proteins containing a sec-dependent signal with the program MEME [57], showed one motif namely, CHxC occurring in several proteins. Forty CHxC proteins were found, the motif showed strong conservation of two Cys residues with no detectable homolog in Phytophthora or Hyaloperonospora species. The high level of sequence divergence from other species is indicative of their specific adaptation in Albugo species. Albugo effector molecules are still not identified. The identification of Albugo effectors and their compatible plant NLR receptors will be a breakthrough, enhancing the scopes of deploying white rust resistance in the susceptible crops with economic importance, like the oilseed Brassicas.

4. PLANT RESPONSE AGAINST OOMYCETES 
EFFECTORS

4.1. Plant Genetic Resistance Architecture

“Genetic resistance” is a phenomenon exercised by plants upon encountering a virulent pathogen or effector molecules with the help of genetic elements to restrict the entry and establishment of pathogens. The resistance mechanism culminated by plants is a result of either expression of disease resistance genes (R genes) or loss or mutation of susceptibility genes (S genes).

Extensive studies have been executed to understand the role of the R gene by overexpressing them in susceptible crop varieties. The R genes convey plant disease resistance against pathogens by producing R proteins and have successfully proved themselves as a promising genetic source for resistance. In the 1940s, a classic “gene-for-gene” theory was established, which states that the outcome of any given plant-pathogen interaction is largely determined by a R gene from the host and the matching avirulence factor (avr) from the pathogen [58]. When both the R gene in the plant host and the cognate avr in the pathogen are present, the plant-pathogen interaction becomes incompatible and the host exhibits full resistance to the pathogen [59]. Guard model elucidates that the R gene products may not directly bind to avirulence gene products, but rather detect alterations in host proteins that are caused by the pathogen gene products [60]. The effectiveness of R gene-mediated resistance was first demonstrated in the early twentieth century by British scientist Rowland Biffen in wheat (Triticum sp.) breeding [61]. The majority of known R proteins are grouped into a few main classes based primarily upon their combination of a limited number of structural motifs. Most R proteins restrict only pathogen races, which express the corresponding effector protein(s) and therefore known as race-specific R proteins. Occasionally, the effective resistance is conferred against multiple races and even different pathogen species. These R proteins are called race-non-specific. The R protein structures of monocotyledonous and dicotyledonous species possess striking similarity implying that fundamental modes of pathogen recognition as well as defense signaling have been preserved through plant evolution and diversification [62]. The majority of plant R genes encode nucleotide-binding site leucine-rich repeat (NLR) type protein which can be further grouped into three subclasses on the basis of the NB-ARC domain. The three monophyletic classes of plant NLRs can be distinguished by their N-terminal domains; toll/interleukin1 receptor (TIR) NLRs (TNLs), Coiled-coil (CC) NLRs (CNLs) and NLRs containing an N terminal RPW8 domain (RNLs) (Fig. 2) [63]. The N-terminal domain precedes an evolutionarily conserved domain with a Nucleotide-Binding site present in Apoptotic protease-activating factor, R proteins, and Caenorhabditis elegans death-4 protein (NB-ARC), followed by highly variable Leucine-rich repeats (LRR) [64]. CNL and RNL genes are found in both monocots and dicots, whereas TNL genes are restricted only to the dicots (Fig. 2 and Table 1) [65]. Alike PRR, variability in the LRR enables to recognize various effector structures [66]. The NB-ARC contains several conserved motifs: P-loop/Walker-A, resistance nucleotide-binding site A (RNBS-A), Kinase-2/Walker-B, RNBS-B, RNBS-C, GLPL, RNBS-D, and MHDV, GLPL and MHDV named after the conserved amino acids present. LRRs are involved in protein-protein interactions and occur in a number of proteins with diverse functions [67, 68]. Domain exchange between LRR of closely related R genes supports their role in pathogen recognition [59]. Variation among R genes occurs mainly in their LRR domain, typically in the solvent-exposed β-strand/β-turn structure within the LRR domain. A comparison of this motif among R gene homologs suggests that the β-strand/β- turn structure has been under diversifying selection [69]. R genes encoding these intracellular receptors were first identified in Arabidopsis through genetic mapping of R genes [70]. Several other receptor-like genes, actively engaged in different modes of defense regulation have been also identified in A. thaliana and Brassica species conferring resistance against A. candida isolates [44].

Fig. (2).

Architecture of the monophyletic group of plant NLRs. The predicted and experimentally documented functions or properties of the individual domains are indicated, as well as particular features and example. TIR – Toll/interleukin-1; CC- Coiled-coil; RPW8- Resistance to powdery mildew n8; NB- Nucleotide binding; ARC1/2- Apaf-1; LRR- Leucine-rich repeat; NLR – Nucleotide-binding and leucine-rich repeat-containing proteins; TNL-TIR-NLR; CNL- CC-NLR;; RNL – RPW8-NLR. (A higher resolution / colour version of this figure is available in the electronic copy of the article).

Table 1.

The table represents NLR and NLR-like gene numbers corresponding to NB-ARC-LRR-encoding genes. The number for TNLs and CNLs correspond to genes encoding either full-length TNLs, CNLs or the NB-ARC-LRR-containing proteins if these can be clearly assigned to one of the NLR types based on their motif composition at the NB-ARC domain. The respective complete genome sizes are also mentioned.

Species	Common Name	Genome Size (Mbp)	NLRs	TNLs	CNLs	References
Arabidopsis thaliana	Thale cress	125	151	94	55	Meyers et al. [65]
Arabidopsis lyrata	Lyre-leaved rock-cress	230	138	103	21	Guo et al. [65]
Brassica rapa	Mustard	485	80	52	28	Mun et al. [65]
B. juncea	Indian mustard	922	NR	NR	NR	Paritosh et al. [99]
B. oleracea	Wild cabbage	630	NR	NR	NR	Paritosh et al. [99]
B. napus	Rapeseed	1130	NR	NR	NR	Paritosh et al. [99]
Carica papaya	Papaya	372	34	6	4	Porter et al. [65]
Cucumis sativus	Cucumber	367	53	11	17	Wan et al. [65]
Glycine max	Soybean	1115	319	116	20	Kang et al. [65]
Medicago truncatula	Barrel medic	186a (500)	270	118	152	Ameline-Torregrosa et al. [65]
Oryza sativa	Rice	466	458	0	274	Li et al. [65]
Sorghum bicolor	Sorghum	760	184	0	130	Li et al. [65]
Solanum tuberosum	Potato	840	371	55	316	Jupe et al. [65]
Vitis vinifera	Wine grape	487	459	97	215	Yang et al. [65]
Zea mays	Maize	2400	95	0	71	Li et al. [65]

In genetic term, S genes can be defined as genes that are involved in disease susceptibility thus contributing positively to the infection progression. Thus, all plant genes that facilitate infection and support compatibility can be considered as S genes. Functional impairment of the S gene by mutation or deletion can alter the ability of the pathogen to cause disease. This may lead to pathogen-specific resistance if the gene is involved in the production of a component required for host penetration or in broad-spectrum resistance if the gene suppresses constitutive defenses. S gene was successfully identified first time in A. thaliana, while exploring resistance against H. parasitica (powdery mildew) [33]. S genes that have been identified as susceptibility factors for colonization by important oomycetes are listed in Table 2. S genes can be categorized into three groups based on the point at which they act during the infection process. These three are those involved in early pathogen establishment, negative regulators of plant immunity, such as the CesA3 gene, which is involved in cellulose synthesis and pathogen sustenance such as metabolite biosynthesis and sugar transport [33].

Table 2.

Cloned resistance (R) and susceptibility (S) genes affecting oomycete-plant interactions.

Oomycete Pathogen	Cloned R Gene(s)	Source of R Gene(s)	Cloned S Gene(s)	Source of S Gene(s)
Albugo candida	RAC1,WRR4A, WRR4B-Col-0, WRR4B-Ws,WRR8, WRR9, WRR12 [27, 30]	Arabidopsis	NR	NR
Hyaloperonospora arabidopsidis	RPP1 (ATR1), RPP2, RPP4, RPP5, RPP7, RPP8, RPP13 (ATR13) [70]	Arabidopsis	AGD5, IOS1, PUB22/23/24 (154, 55), SON1, EDR2, SNI1,Cdd1, DMR1, RSP1/2, PMR4, DMR6 [70]	Arabidopsis
Peronospora manshurica	Rpm [33]	Soybean	MPK4 [33]	Soybean
Phytophthora cinnamomi	TIR1 [33]	Arabidopsis	NR	NR
Phytophthora infestans	R1, R2 (AVR2), R3a (AVR3), R3b (AVR3b), R4 (AVR4), R6 and R7, R10 and R11, RB/Rpi-Blb1 (AVR-Blb1/IPI-O1), Rpi-Blb2, Ph-3, Rpi-vnt1, Rpi-blb3, Rpi-abpt [33]	Potato	StREM1.3 and N. benthamiana REM1.3 orthologs [33]	N. benthamiana
Phytophthora palmivora	NR	NR	RAM2, LATD [33]	Medicago
Phytophthora sojae	Rps1d (AVR1d), Rps1b (AVR1b) [33]	Soybean	NR	NR
Plasmopara viticola	Grape Rpv1 and Rpv2, Rpv3 (avrRpv3), Rpv10 [33]	Grape	NR	NR

(NR - Not Reported).

4.2. Manifestation of Oomycetes Resistance in Plants by NLR Signaling (with Reference to A. candida)

Plants can sense diverse extracellular oomycete-derived patterns. Oomycete patterns are derived from the pathogen’s cell wall or membrane, whereas others are secreted to the extracellular environment before being detected by the plant's immune system.

As we have already discussed above, oomycete pathogens secrete effectors during invasion and propagation into the plant cell. The cytoplasmic nucleotide-binding and leucine-rich repeat receptors (NLR) present in plants detect oomycete effectors or their activity, thus triggering the immunity (ETI). Many effector–NLR interactions have also been studied in Arabidopsis [71]. Several key plant defense regulators, downstream of NLRs have been identified which includes ENHANCED DISEASE SUSCEPTIBILITY 1 (EDS1), and two NLR helpers - ADR1 and NGR1. EDS1 is a lipase-like protein that is required for the functioning of TNLs, leading to oxidative stress signaling and as a positive regulator as well as basal resistance to virulent pathogens [72]. EDS1 is well reported to be involved in the functioning of TNL genes such as RPP1, RPP4, RPP5, and WRR4. NLR helper proteins play a very significant role in the NLR downstream signaling. ADR1 and NGR1 are two RNL proteins, which have been recently identified as NLR helper proteins and transmit effector immunity in the host infected by A. candida and other notable oomycetes. ADR1 helps in the signaling of both TNLs and CNLs receptors along with the regulation of the Salicyclic acid (SA) dependent pathways. NRG1 is found to help only TNLs during the downstream signaling coupled with regulation of HR and other defense responses. Additionally, NRG1 has also shown complete resistance to A. candida (Fig. 3). On the basis of the previous findings by Cevik et al. (2019) [30] and Caster et al. (2019), we understand that direct or indirect detection of the A. candida effector molecules by a flowering plant harboring NLRs leads the NLRs to undergo a conformational change from a condensed, ADP-bound state to an open ATP-bound state with exposed N-terminal domains for the initiation of downstream signaling [66, 73-75].

Fig. (3).

Schematic representation of PTI and ETI in plants against A. candida. The recognition of the PAMPs of A. candida by PRR leads to PTI. Specific activation of TNL & CNL receptor by A. candida effectors triggers ETI. Few TNL genes are identified, for example, RAC1, WRR4A, WRR4B, WRR8, WRR9, and WRR12 reported in Aradopsis [30]; CNL genes, for example, BjuWRR1 in Donskaja IV, is an east European variety of B. juncea [77]. All the TNLs signal via NRG1 and some can also signal via ADR1. CNLs signal via ADR1 and partially via NRG1. ADR1 and NRG1 are conserved within both monocots and dicots suggesting this model can be applied for all the flowering plants. WRR4A gene has been proved to signal via NRG1 [80]. Rest TNL genes - RAC1, WRR4B, WRR8, WRR9, WRR12 are hypothsised to signal via NRG1 and ARD1, whereas the CNL gene(s) – BjuWRR1 is hypothesized to signal ADR1. (A higher resolution / colour version of this figure is available in the electronic copy of the article).

The activated NLRs along with the other proteins like EDS1 reported in B. juncea, B. napus, and A. thaliana initiate a downstream defense-signaling pathway. Furthermore, they are transmitted by ADR1 intracellular receptor protein in the case of both TNLs and CNLs and NRG1 in the case of only TNLs (Fig. 3). We can conclude suggesting that TNL, CNL and RNL proteins, on activation, trigger a complex network of responses, including gene induction, the production of reactive oxygen species and salicylic acid, transcriptional reprogramming and finally lead to cell death phenomenon ‘Hypersensitive Response (HR)’, resulting in host resistance.

5. NOVEL APPROACHES TO TRANSFER WHITE RUST RESISTANCE IN BRASSICA JUNCEA

The Indian gene pool lines of B. juncea are highly susceptible to the A. candida [76]. Contrary to the success in the production of oilseeds through conventional techniques, the development of oomycetes resistant brassicas has been limited. This drives an urgent need to blend the conventional, unconventional and frontier technologies for crop protection. Some accessions of A. thaliana and the east European mustard germplasm harbor resistance against Ac2VRR as well as other A. candida races [30, 76]. The non-host resistance factors present the resistant plants that can be of significant interest to introduce WRR in the susceptible commercial varieties of B. juncea. Studies report that the non-host resistance present in A. thaliana is majorly due to the presence of R genes [27]. Therefore, the transfer of A. candida effector compatible R genes through genetic engineering gives us high scope to deploy durable white rust resistance in the susceptible commercial varieties of B. juncea (Fig. 4).

Fig. (4).

Hypothetical figure for application of novel genomic approaches to transfer white rust resistance in B. juncea. A) Transfer of R genes in B. juncea by genetic transformation followed by overexpression leads to deployment of resistance; B) Inhibition of the target S genes in B. juncea by introduction of antisense oligonucleotides by genetic transformation leading to resistance; C) Host Induced Gene Silencing (HIGS) mediated by the siRNA in the pathogen, while invasion of the genetically modified siRNA synthesizing transgenic B. juncea; D) The targeted knocking in R genes and knocking out of S genes in the susceptible B. juncea varieties. (A higher resolution / colour version of this figure is available in the electronic copy of the article).

The genome of B. juncea variety Tumida and B. juncea variety Varuna has been sequenced [77]. Further analysis of the sequences will enable us to identify S genes responsible for WR infection in the host tissue. Genomic approaches like RNA interference and CRISPR-Cas will help us to silence or knock out the pathogen Avr genes and host S genes respectively.

Here, we have tried to brief the novel genomic approaches which can help us to develop WRR Indian mustard (Fig. 4).

5.1. Deploying Resistance Genes for Broad-Spectrum and Durable Resistance

Plant disease resistance genes (R genes) play a key role in recognizing proteins expressed by specific avirulence (Avr) genes of pathogens encoding the effector molecules. The discovery of R genes (including their homologs and analogs) opened interesting possibilities for controlling plant diseases caused by several pathogens. However, due to high selection operated by pathogens along with environmental pressure, several crop plants have lost their specificity, broad-spectrum or durability of resistance. On the contrary, the advances in plant genome sequencing and identification of numerous R genes have provided new insights into the acquisition of resistance to pathogens in host plants through transgenesis, or gene editing. The main class of R genes consists of a nucleotide-binding domain (NB) and a leucine-rich

repeat (LRR) domain(s) and often referred to NLR/NB-LRR as discussed already in section 4.1. The NLR family is massively expanded in several plant species. The number of NLR genes in flowering plants is largely variable suggesting species-specific mechanisms in NLR genes expansion and/or contraction (Table 1). This variability can be explained by three species in the Brassicaceae family; A. thaliana, A. lyrata, and B. rapa, which have 151, 138, and 80 full-length NLRs, respectively. The first NLR gene, RPS2, was cloned from Arabidopsis conferring resistance against Pseudomonas bacteria, expressing the AvrRpt2 effector molecule [70].

Scientists across the globe are working to explore R genes conferring WRR. Several R genes conferring WRR against different races of A. candida have been mapped in A. thaliana, B. rapa, B. napus, B. carinata and B. juncea [30]. Among the identified WRR genes, RAC1(Arabidopsis accession Ksk-1) and WRR4 (Arabidopsis accession Col-0) were identified and cloned from A. thaliana [30]. WRR4 gene was overexpressed into susceptible cultivar of B. juncea and B. napus by genetic transformation and was reported to confer significant levels of resistance against Ac2VRR and Ac7V [45]. WRR4 gene has been studied to arrest the development of the pathogen in the plant epidermal cell, which shows hypersensitivity response [30]. Recently, a few more genes namely WRR4BCol-0, WRR8Sf-2, WRR9Hi-0, WRR12 have been identified in Arabidopsis accessions conferring resistance to WR [30]. All these genes belong to the TNL subfamily (Tables 1 and 2). Few candidate genes have also been mapped in two east European B. juncea gene pool lines Heera and Donskaja-IV, and one Chinese B. juncea variety Tumida conferring resistance against Indian isolate AcB1 (Albugo candida Bharatpur-1) [76]. Gene BjuWRR1 from Donskaja IV, encoding a CNL domain-containing protein has been identified to confer WRR. The transgenic B. juncea variety Varuna constitutively expressing BjuWRR1 conferred WRR to various isolates of A. candida [77].

5.2. Enriching the Known Repertoire of Immune Receptors

Plants possess immune receptors that recognize pathogens and trigger cellular defense responses [13]. Discovery of novel immune receptors recognizing major virulence factors will enrich the repertoire of known immune receptor genes that may be deployed in the crop genome for better performance in the field [78]. Once a collection of germplasm exhibiting various degrees of resistance (non-host resistance) to a particular pathogen strain is identified, a comparative genomic tool such as resistance gene enrichment sequencing (RenSeq) can be applied to identify genomic variants in NLR genes that are linked to disease phenotypes [79]. This promotes the cloning of new NLR genes and their potential deployment in crop protection through genetic engineering. RenSeq was successfully applied in the accelerated identification of NLRs for example, anti-P. infestans NLR gene Rpi-amr3i from Solanum americanum, anti-potato-late-blight NLR gene Rpi-ver1, stem rust NLR genes, Sr22 and Sr45, which confer resistance to commercially important races of the stem rust pathogen from a mutagenized hexaploid bread wheat population [79].

Further, focusing on white rust disease, host/race specificity of A. candida is determined by the NLR repertoire of the host plant and the cognate effectors of the pathogen race, rather than host compatibility factors. Therefore, some of the NLRs recognizing specific races or multiple races are maintained in different Brassicaceae species. This, in turn, provides an excellent resource to identify WRR genes for different Brassica species.

RenSeq enabled to identify three WRR genes (WRR4BCol-0, WRR8Sf-2, and WRR9Hi-0) against B. juncea- infecting A. candida race 2 (Ac2V), and a gene, WRR12 (SOC3), conferring NHR to AcBoT, in addition to the previously identified broad-spectrum resistance gene WRR4ACol-0 (previously known as WRR4) [13, 30]. The distribution and sequence variation of WRR4A, WRR4B, WRR8, WRR9, and WRR12 genes was determined while sequencing the bulk DNA of different accessions of Arabidopsis using SMRT RenSeq [30]. The sequences of the WRR alleles from each A. thaliana accession were identified by BLASTn against the SMRT RenSeq assemblies [80]. BLASTn hits with less than 95% identity, are not considered the alleles of the query gene [30]. Although future field experiments are required to evaluate the transfer potential of these newly identified NLR genes in the host. These lab studies confer that RenSeq is a powerful tool to rapidly identify novel NLR genes (Tables 1 and 2).

5.3. RNA Interference Technology

RNA interference (RNAi) is a potential tool for silencing genes in a broad range of organisms. It involves the expression of a double‐stranded RNA (dsRNA) or small interfering RNA (siRNAs) that trigger the degradation of target mRNA sequences by post‐transcriptional gene silencing (PTGS). It is not so much successful in oomycetes pathogen alike viruses and fungus. Studies have reported that the endogenous miRNAs and siRNAs present in plants, regulate the expression of some NLR genes (Li et al., 2012; Prasad et al., 2012). This was validated by demonstrating the function of the diverse endogenous miR482 family in tomato targeting NLR transcripts in a sequence‐specific manner and induced subsequent cleavage of the transcripts. This paves the way for the application of RNAi through the transfer of artificial miRNAs and siRNA in the plants for suppression of S genes and for Avr gene suppression in the pathogen. Genetic transformation helps to transfer the siRNA and miRNA constructs in the target plant or pathogen. There are several reports of gene silencing in plants through RNAi [79]. Polyploidy in plant challenges the targeted silencing of specific homolog(s) of a multigene family RNAi [81, 82]. Oilseed crop B. juncea (AABB) is an amphidiploid which is expected to have 4-6 copies of every Arabidopsis gene ortholog [82]. A recent study in polyploid B. juncea suggests that designing intron-spliced hairpin RNAi constructs with optimally expressed gene targets and promoters can lead to efficient silencing, especially when a specific member of a multigene family in a polyploid crop is targeted [83].

S genes render disease susceptibility in plants thus contributing positively to the infection process. On the basis of their working mechanism, S genes have been classified into three classes as discussed in section 4.1. Silencing S genes involved in pathogenesis has proven to suppress few significant bioprocesses supporting pathogenesis leading to enhanced disease resistance in plants (Fig. 4). The bioprocess can be biosynthesis of wax, cellulose, metabolites and sugar transport [84]. This strategy has been very successfully applied to plant viruses [85]. Silencing of S genes through RNAi has also shown its success in Arabidopsis (against downy mildew) as well as significant crop plants like potato (late blight resistance, Botrytis blight), tomato (powdery mildew) as detailed in Table 2. The orthologs of the S genes whose silencing has conferred resistance to these above-mentioned oomycetes diseases can also be the target candidate genes for attaining WRR in the host.

RNAi also gives us a platform to target the host invading pathogen. The genes identified in important bioprocesses of the oomycetes pathogen, for example, chitin synthase gene required for chitin synthesis (an important component of cell wall), carbohydrate myco-laminarin gene (involved in energy storage) as well as avirulent genes are an ideal target for silencing [86]. Gene silencing has not yet been reported in A candida but has been applied to several Phytophthora species to generate strains deficient in particular gene products. This can be a thrust area for future studies. The transformation of P. infestans with sense, antisense, and promoterless constructs of the pathogen endogenous gene, for example, inf1 gene, leads to the silencing of the inf1 gene in as many as 20% of the transformants [86].

RNAi based Host Induced Gene Silencing (HIGS) is also a potential strategy for silencing the pathogen Avr genes of the crop infecting pathogen. Avr genes encoding the effector molecules have been proved to be ideal targets to attenuate the pathogen during the invasion of the host cell. As reported, HIGS has been successful in limiting various phytopathogens such as insects, nematodes, viruses (Virus-Induced Gene Silencing), and fungus [79]. Interestingly, effectors from oomycetes have been identified that function as suppressors of RNA silencing [79]. Gene knockouts are easier in fungus as compared to oomycetes [86]. Oomycetes are typically diploid during their asexual stages, which means for gene disruption both the copies of the gene have to be targeted. In oomycetes, most studies that involve gene silencing have been conducted in P. infestans [80]. Classical gene disruption approaches have not been so much successful with oomycetes, since homologous recombination occurs at very low levels. Therefore, attempts at targeted gene knockout have dominated on the gene-silencing strategies that have proven successful in plants

The advancement in B. juncea and A. candida genome sequencing will enable us to identify the S genes and Avr genes present in it providing future scopes for the attainment of WRR in Indian mustard.

5.4. Scope of CRISPR for Developing White Rust 
Resistance in Plant

In recent years, the bacterially derived clustered regularly interspaced short palindromic repeats (CRISPR)-CRISPR-associated (Cas) technology has proven to be a promising approach for engineering resistance to plant viruses [87]. CRISPR-Cas site-specific nucleases evolved as components of prokaryotic immunity against viruses, and are widely deployed as tools to deploy operator-specified nucleotide sequence changes in genomes of interest. It can be used to precisely modify the genome of any organism including plants to achieve the desired trait. In the past decade, CRISPR emerged as a cheap, fast, precise and highly efficient tool as compared to other genome editing tools such as zinc finger nucleases (ZFNs) and transcriptional activator-like effector nucleases (TALENs), even at the multiplex level to edit genomes and create genetic variations quickly addressing emerging challenges in agriculture [84]. The CRISPR/Cas9 is becoming a user-friendly technique for development of non-transgenic genome edited crop plants to counteract the phytopathogen menace ensuring future food security of the increasing population in tropical countries.

In plants, it enables to knock-out the S genes by targeted mutagenesis and knock-in the R genes. Random mutagenesis can be utilized to generate polymorphism and characterize genes in a forward genetic approach (Fig. 4). However, the large genome size and high copy number in polyploidy crops possess several challenges in site-directed mutagenesis for knocking-out of multiple genes with high homology. In 2010, Birch et al. [38] showed that the transgene-silencing is promotor sequence-specific, which makes it obligatory to choose diverse and proficient promoters to trim down the silencing effects [81]. Complete annotated genomes will enable us to design sgRNAs to target genes in the crop. The requirement of a large number of mutants to study the multiple forms of alleles present in polyploid is another fundamental limitation [81]. As we have discussed, S genes are responsible for susceptibility in plants. It has been demonstrated to be efficient for gene disruption in many plant species, including Arabidopsis, Nicotiana benthamiana, N. tabacum, rice, wheat, maize, sorghum, tomato, potato, sweet orange, poplar, and liverwort [88]. Recent studies report, several examples of CRISPR-mediated S gene disruption resulting in resistant plants [80]. MLO, a dominant S gene that confers broad-spectrum susceptibility to powdery mildew in diverse plants from monocots to dicots was knocked out by CRISPR. The raised mlo2-mlo6-mlo12 triple mutant was fully resistant to an Arabidopsis-compatible powdery mildew strain [80]. CRISPR was used successfully to knock-out MLO genes in wheat and tomato inculcating resistance to powdery mildew. Lab study demonstrated that overexpression of SpCas9 and guide RNAs in N. benthamiana, Arabidopsis (Arabidopsis thaliana), tomato conferred resistance to the geminivirus family member Beet severe curly top virus [87]. CRISPR-Cas has also been successful in engineering resistance to RNA viruses, which comprise the most known plant viral pathogens [88]. For example, stable expression of the RNA-targeting nuclease Cas13a and the corresponding guide RNA in N. benthamiana conferred resistance to the RNA virus Turnip mosaic virus. Using Cas13a to target viral RNA substrates does not induce DNA breakage and thus would not introduce undesired off-target mutations to the host genome (Abudayyeh et al., 2017). Similarly, FnCas9 has been used to engineer resistance against RNA viruses Cucumber mosaic virus and Tobacco mosaic virus in N. benthamiana and Arabidopsis [84].

In Brassica, only a few successful instances of genome editing have been reported [88]. One of these cases demonstrated the GA4 gene knockout in a doubled-haploid genotype AG DH1012 (a broccoli-like Brassica) from the Brassica oleracea var. alboglabra (A12DHd) × B. oleracea var. italica (Green Duke GDDH33) mapping population [89]. The BoPDS gene, the self-incompatibility gene BoSRK3, the BoMS1 gene associated with male sterility, and some of their paralogous genes were successfully knocked out using an array of sgRNA-tRNA units designed to express more than one sgRNA and showed obvious mutant phenotypes in the T0 generation. In addition, the CRISPR/Cas9 system can target multiple sites or multiple genes in a single transformation event and produce homozygous knockouts, even in the T0 generation, as reported in other species [88]. This approach enables us to develop resistant crops in a few months, with mutation undistinguishable from naturally occurring mutation.

CRISPR has helped to identify the function of two NLR helper genes: NRG1 and ADR1 in Arabidopsis and N. benthamiana during A. candida effector-triggered downstream signaling [84]. The two genes were studied to redundantly understand the full function of CC-NLR Rx2 and TIR-NLR pair RRS1/RPS4 [85]. The generation of null alleles via CRISPR is quick and simple nowadays, facilitating the investigation of gene function.

Reviewed literatures indicate that homology-directed recombination (HDR) mediated repair of CRISPR-induced double-strand break (DSB) can result in gene replacement or insertion in plant and animal. Gene targeting via HDR occurs through homologous recombination between a template and genomic DNA at very low frequency and drastically increases when DSB occurs [84]. Although CRISPR-mediated gene targeting is a routine assay in animals, it remains presently a challenge in plants. Stable gene targeting was achieved in Arabidopsis and tomato in 6% to 25% of transformed plants [90]. Theoretically, CRISPR could be employed to introgress an R-gene in a susceptible line quickly and with no other DNA introgression as compared to classic breeding. It is yet to be demonstrated. In a small span of five years, CRISPR has been deployed to develop significant resistance in plants via RNA virus targeting and S gene knock-out. If gene targeting is one of the current challenges, we can expect to witness much more CRISPR-derived applications in the coming years, for plant disease resistance and beyond.

However, the lack of genetic data for B. juncea impedes the rapid cloning of WRR genes from this species. The advances in B. juncea sequencing and optimization of CRISPR in the polyploid B. juncea will enable us to integrate the WRR genes in the susceptible varieties of B. juncea, deploying durable resistance to white rust disease. CRISPR-based genome editing techniques will facilitate the study of genes and proteins and will be beneficial for both basic and applied plant science (Fig. 4).

Conclusion and future directions

A. candida, a homothallic biotrophic oomycete, can form hybrids between genetically distinct isolates via co-culture recombining genetic variability in A. candida populations [15]. Comparative genomics has revealed an extensive genetic exchange between races of A. candida and this genetic exchange could result in races with novel repertoires of effector alleles that, in turn, might enable colonization of new hosts. Therefore, understanding the underlying mechanism of NHR in different Brassica species could help us in breeding for resistance to A. candida. Host/race specificity of A. candida is determined by the NLR repertoire (encoded by R genes) of the host plant and the recognized effectors of the pathogen race rather than host Identifying R and S genes in B. juncea as well as other species of Brassica is must for the deployment of resistance to white rust. Rapid identification of more R genes with advanced genomic tools such as RenSeq should be practiced to enrich the NLR protein repertoire enhancing the prospects of varied pathogen recognition and the manifestation of ETI. For example, a study reported that ResSeq enabled the identification of the R genes namely WRR4B Col-0, WRR4B Col-0, WRR8 Sf, WRR9-Hi in Arabidopsis deploying resistance to B. juncea-infecting race Ac2V and WRR12 in Arabidopsis deploying resistance to B. oleracea-infecting race AcBoT encoding NLR [30]. Efforts to identify R genes have also been assisted by the development and application of differential host-pathogen pathotype/race/isolate combinations. Further improvements can be improvised by establishing freely-available universal host differential sets for each pathogen, differential sets that constitute the host resistances, and pathogen sub-specific variation worldwide. One of the approaches that could be included is an international collaborative network building a world database for each pathogen. The initiative established at the Brassica 2016 conference to resolve the Blackleg nomenclature issue is a remarkable step for this approach [91]. Since dynamic changes can happen with either of the host cultivars and pathogens in terms of cultivar resistance breakdowns and/or pathogens losing/gaining in terms of their pathogenicity/virulence over time, every universal database set up must be regularly updated and kept current. For example, the international DivSeek database, for phenotypic data that can be linked with genotypic data (for seeds that are stored in seed bank), that was launched in 2015 could be utilized as a model to set up the databases needed. Successful phytopathogens often evade detection by host R genes [92]. Thus, disease resistance conferred by a single R gene often fails to provide durability in the field as pathogens can evolve to evade recognition by mutating the corresponding Avr gene. For improved durability and to broaden the resistance spectrum, multiple WRR genes can be introduced simultaneously, which is commonly known as stacking [93]. Three late blight resistance genes have been stacked and successfully transformed in highland potato varieties deploying durable resistance to late blight. Functional impairment of S genes in the host plant is also a favorable approach for sustaining disease resistance in Indian mustard as well as other economically significant crops. RNAi interference and CRISPR-Cas technology are two promising approaches which can knock out the S genes impairing the disease manifestation in the host plant. Future efforts in improving CRISPR-Cas for anti-oomycete resistance may also be focused on establishing an in-plant adaptive immunity by exploiting the spacer acquisition machinery in the CRISPR-Cas adaptive immune system in prokaryotes [94].

Along with the genomic approaches, the plant and seed microbiome of oilseed Brassica is an interesting aspect and should be explored to develop a resistance strategy against Albugo infection. Plant microbiome includes a microbial community that typically interacts extensively with a plant [95-97]. The plant microbiome can survive either inside or outside of plant tissues including seed and phyllosphere, performing various plant beneficial activities such as biocontrol potential against phytopathogens and promotion of plant growth. The seed microbiome of oilseed rape, is cultivar-specific [95], posing a strong influence on its interaction pattern with pathogens. Therefore considered as an interesting biomarker for breeding strategies [95]. We recommend that microbiome studies should be incorporated into breeding programs for sustainable agriculture practices. We also believe that the assessment of seed and phyllosphere microbiomes combined with network analysis may open new opportunities for the targeted selection of biocontrol strains for a given host plant at the cultivar level. These novel insights into plant and seed microbiome structure will enable the development of next-generation strategies combining both biocontrol and breeding approaches to address world agricultural challenges [98, 99].

Furthermore, the advancement in the Brassica-Albugo genomics will help us to identify more R, S and A. candida avirulent/pathogenic genes unraveling the underlying infection and non-host resistance mechanism, thus enabling us to develop white rust resistant Indian mustard.

Consent for Publication

Not applicable.

Funding

This study is supported by the Department of Biotechnology (DBT), Government of India under the Indo - UK project, Grant no. BT/IN/Indo-UK/CGAT/12/DP/2014-2015 dated 8^th May 2015 (PI/CoPI – Dr. Abha Agnihotri and Dr. Pankaj Sharma).

CONFLICT OF INTEREST

The authors declare no conflict of interest, financial or otherwise.