Understanding R Gene Evolution in Brassica

Abstract

Brassica crop diseases caused by various pathogens, including viruses, bacteria, fungi and oomycetes, have devastating effects on the plants, leading to significant yield loss. This effect is worsened by the impact of climate change and the pressure to increase cultivation worldwide to feed the burgeoning population. As such, managing Brassica diseases has become a challenge demanding a rapid solution. In this review, we provide a detailed introduction of the plant immune system, discuss the evolutionary pattern of both dominant and recessive disease resistance (R) genes in Brassica and discuss the role of epigenetics in R gene evolution. Reviewing the current findings of how R genes evolve in Brassica spp. provides further insight for the development of creative ideas for crop improvement in relation to breeding sustainable, high quality, disease-resistant Brassica crops.

1. Introduction

The Brassica genus (family Brassicaceae) contains economically important crop species that are widely grown throughout the world in the Americas, Asia, Europe and Oceania [1].There are six Brassica species that form the main vegetable and oilseed food crops, including three diploid members, namely B. rapa (rapeseed, AA genome, n = 10), B. nigra (black mustard, BB genome, n = 8) and B. oleracea (cole crops, CC genome, n = 9); there are also three allotetraploid members which were formed from pairwise hybridisation of those three diploid members, namely B. juncea (Indian mustard, AABB genome, n = 18), B. napus (canola, AACC genome, n = 19) and B. carinata (Ethiopian mustard, BBCC genome, n = 17), of which the relationship can be described as the triangle of U [2].

Brassicas, which are closely related to Arabidopsis (tribe Arabideae), are a member of the tribe Brassiceae within the family Brassicaceae, constituting nearly 50% of the entire 3740 species in the family [3], and are represented by large morphological diversity [4]. After the split of the Arabideae and Brassiceae tribes 5–9 million years ago, whole-genome triplication (WGT) of the hexaploid Brassica ancestor occurred, leading to massive chromosomal rearrangements, with re-construction to a more stabilised diploid Brassica species belonging to the Brassiceae tribe through many rounds of polyploidisation [5,6,7]. The benefit of the WGT event unique to the Brassica lineage is the expansion of genes related to abiotic and biotic stress adaptability along with plant hormonal networks [5].

One of the challenges in sustaining Brassica crop yields is biotic stress. Various pathogens, including viruses, bacteria, fungi and oomycetes, can infect Brassica crops, causing major economic losses. The main diseases of Brassica are caused by the bacterium Xanthomonas campestris (black rot), turnip mosaic potyvirus (TuMV), the oomycetes Albugo candida (white rust) and Hyaloperonospora parasitica (downy mildew), the protist Plasmodiophora brassicae (clubroot) and the fungi Alternaria brassicae (Alternaria blight), Erysiphe cruciferarum (powdery mildew), Fusarium oxysporum (Fusarium wilt), Leptosphaeria maculans (blackleg), Neopseudocercosporella capsellae (white leaf spot), Sclerotinia sclerotiorum (Sclerotinia stem rot) and Verticillium longisporum (Verticillium wilt), amongst others.

To survive and multiply in the natural ecosystem, plants have developed an innate immune system to defend themselves against pathogen invasion. Substantial research has shown that plants are equipped with a two-tiered innate immune system; pathogen-associated molecular pattern (PAMP)-triggered immunity (PTI) and effector-triggered immunity (ETI), in terms of the types of pathogen molecules interacting with the host receptors that activate the immune system [8]. Responsible for non-host resistance, PTI is triggered when cell surface-localised or transmembrane pattern recognition receptors (PRRs) recognise extracellular pathogen/microbial-associated molecular patterns (PAMPs/MAMPs). During the co-evolution of the host–pathogen process, pathogens have developed pathways to evade PTI and secrete effectors that enter plant cytoplasm. ETI is then activated by the recognition of these specific pathogen effectors, or avirulence (Avr) proteins, via corresponding intracellular receptors, leading to a complete resistance. The receptors are mainly encoded by a large resistance gene family featuring nucleotide-binding site leucine-rich repeats (NLRs). Due to the specific recognition involved, this host-mediated resistance is also known as gene-for-gene resistance [9,10,11,12]. Proteins encoded by the resistance (R) genes are termed R proteins, referring to PRRs and NLRs. PRRs are mainly from the receptor-like kinase (RLK) and receptor-like protein (RLP) gene families [13]. NLRs can be further divided into two subtypes according to domains at their N-terminus. With the presence of a Toll/interleukin-1 receptor (TIR) domain, they are classified as TNLs, while with the presence of a coiled coil (CC) domain, or in a rather rare situation, with the appearance of resistance to the Powdery Mildew 8 (RPW8) domain, they are classified as nTNLs, termed as CNLs or RNLs, respectively [14,15,16,17].

Though PTI- and ETI-triggering signals lead to several shared cellular responses, how PTI and ETI interplay spatially to arrest pathogens is poorly understood [18]. Recent groundbreaking studies in the model plant Arabidopsis have revealed a revised immunity model in which PTI and ETI undertake their roles with mutual potentiation [18,19,20]. According to two independent studies in Arabidopsis, PTI is required for the effective complete resistance of ETI; in other words, the activation of surface receptors can effectively enhance the intracellular-receptor-dependent hypersensitive cell death response (HR) and impede pathogens from further propagation [19,20]. Furthermore, the activation of NLRs in ETI, such as TIR signalling, augments the accumulation of PRR signalling components and boosts plant defence during PTI by upregulating the transcripts and protein levels in Arabidopsis lines induced by the PTI elicitor flg22/nlp20 [18]. The mutual enhancement suggests that breeding for crops with accumulated R genes not only strengthens host-mediated resistance but also supports non-host resistance. This synergistic work between PTI and ETI demonstrates the complexity of the plant immune system, which necessitates taking this synergy into consideration in the study of plant resistance gene evolution (Figure 1).

In this review, we expound the sources and variations of Brassica resistance genes, highlight forces of evolution on disease resistance genes in Brassica species and discuss how these forces shape immunity in the plant, describe the cutting-edge and less studied perspectives related to the study of R gene evolution and review the latest technologies applied in this field.

2. Evolutionary Origin of R Genes

2.1. Polyploid Ancestry

On the evolutionary origin of the Brassica species, it has been suggested that ancient B. rapa and B. oleracea were domesticated separately in China and Europe until both species were introduced and crossed with each other during the opening of Silk Road that links China and the Mediterranean, with further backcrossing creating the diverse morphotypes that we see today [21]. More recently, a genotyping-by-sequencing (GBS) study on the origin of B. rapa, the largest diversity study to date on representatives of domesticated and weedy B. rapa, pointed towards a true wild relative lineage or highly admixed feral lineage from the Caucasus mountains situated at the intersection between Europe and Asia, with a domestication event of turnips and/or oilseed types in Central Asia happening 3000–5000 years ago, followed by distribution across Europe and East Asia [22]. In B. oleracea, an updated RNA-seq study proposed the origin of B. oleracea being a monophyletic ancestor coming from the Eastern Mediterranean and the likely progenitors being B. cretica and B. hilarionis, with events of gene flow happening between these wild species (and other wild relatives such as B. incana, B. montana, etc.) and cultivated species throughout the domestication process [23]. The origin of B. juncea was reported to be in West Asia 8000 to 14,000 years ago via natural interspecific hybridisation with subsequent independent domestication events happening near Central Asia, the Indian subcontinent and East Asian regions, giving rise to different morphotypes of mustards [24].

When investigating the relationship between the Brassica species, comparative genomics studies, using mitochondrial and chloroplast DNA, revealed that B. napus (AACC) is genetically closer to B. rapa (AA), B. juncea (AABB) and B. oleracea (CC), while B. nigra (BB) and B. carinata (BBCC) are more diverged from these Brassica species but show a closer relationship with Sinapsis arvensis, a relative of the Brassica species [25]. Whole-genome re-sequencing revealed that the A subgenome of B. napus originated from European turnip B. rapa and the C subgenome from the ancestor of B. oleracea, but it is not known which B. rapa type—the wild or feral type—and the form of the B. oleracea ancestor remains unknown [26]. Nevertheless, gathering all of this information on the historical origin of the Brassica genomes provides context on how disease resistance genes evolve across time through polyploidisation and crop-domestication events.

2.2. Disease Resistance Genes from Introgression Lines

Due to the high compatibility of genomes between members of the Brassica species, R genes can be introgressed through interspecific hybridisation, thereby creating new polyploids with enhanced disease resistance. Aside from producing introgression lines of oilseed B. napus using closely related species or wild relatives to obtain novel disease resistance genes, for instance against blackleg [27,28,29], clubroot [30,31] and Sclerotinia stem rot [32,33], stable allohexaploids with rich sources of R genes have now been created using improved cytogenetics techniques. For instance, interspecific triploid hybrids (ABC genomes) resulting from a cross between B. nigra (B genome) and B. napus (AC genomes) show promising introgression of blackleg resistance genes from the B genome into the A or C genome [34,35,36]. The application of Brassica genus-wide pangenomics (ABC genomes) including related genomes has enabled us to distinguish the donor from the recipient genomes in introgression lines and study the signatures of these hybridised genome patterns [37].

2.3. Studying Disease Resistance Genes from Close Relatives of Brassica

The close genomic relationship between Arabidopsis and Brassica species allows us to discover important Brassica disease resistance genes using the Arabidopsis plant system to look for R gene homologs. This was recently accomplished for Alternaria blight disease: the disease resistance genes were identified in A. thaliana using publicly available microarray datasets and mapped onto the B. rapa genome to identify the homologous genes in B. rapa [38]. The Arabidopsis-Pseudomonas syringae pathosystem shows that RLKs and RLPs are involved in ETI immunity [20], while the Arabidopsis-Albugo candida (white rust) pathosystem revealed the involvement of NLRs [39]. Homologous genes of Fusarium wilt resistance (caused by the fungal pathogen Fusarium oxysporum) in radish, Raphanus sativus, were identified in A. thaliana, B. rapa and B. oleracea genomes [40]. The updated BRAD database, with the expansion of the collection of Brassicaceae genomes, has greatly enhanced our interrogation of R gene homologs [41]. Although many R gene homologs can be inferred in Arabidopsis for many of the Brassica diseases, the predicted race-specific candidate R genes must still be validated through rigorous gene functionality testing to ensure durable resistance in Brassica. With the pan-NLRome available for A. thaliana and the proposed pan-RGAome in the Brassica genus, species-wide R gene homologs can be identified [42,43].

2.4. Structural Variation of Brassica Resistance Genes

A large number of NLR genes against the three major diseases of Brassica—blackleg, clubroot and Sclerotinia stem rot—were found in a selective sweep region, implying strong selection pressure associated with the domestication process and promoting quick diversification of the R genes in this region [44]. The location of R genes in these highly variable regions in Brassica has warranted a search for R gene candidates using a pangenome approach, revealing signification patterns of SNPs and PAVs (presence/absence variation) influencing the R gene diversity in B. napus [45,46] and B. oleracea [47]. A pangenome is developed from a compilation of genome assemblies coming from various Brassica ecotype representatives, and by using sophisticated algorithms, the number and type of core vs. dispensable genes can be uncovered. For example, through the study of the pangenomes of B. napus and its progenitors B. oleracea and B. rapa, it was discovered that gene loss events following polyploidisation in B. napus were linked with homoeologous recombination, differentiating them from gene loss events in B. rapa and B. oleracea linked with transposable elements [48]. Supporting these findings is the in-silico sequence comparison of the NBS-encoding genes in these three Brassica genomes, showing gene expansion and loss through non-homologous recombination [49].

The implementation of genomics and bioinformatics tools in the study of R genes in Brassica species, including members within the Brassicaceae, has greatly increased our understanding about how these genes evolved [17]. Common QTL co-located with NBS genes against blackleg, clubroot and Sclerotinia stem rot have been found in B. napus [49], suggesting multiple disease-resistance effects. Tandem and segmental duplications of the NBS gene family are commonly found in polyploids and diploids such as radish (Raphanus sativus) [40]. Examples of tandemly duplicated TNL genes responsible for Plasmodiophora brassicae, or clubroot, resistance are Crr3T^sc in B. napus and the Cra/Crb/CRb^kato locus in B. rapa [50,51]. Genomic regions associated with quantitative resistance to blackleg were found to be duplicated in B. napus [52], with their density bias being towards the A subgenome, with higher collinearity at the homoeologous loci than paralogous loci [53]. The recently cloned qualitative blackleg R genes in B. napus—Rlm4, Rlm7 and Rlm9—were identified as allelic variants on chromosome A07 [54,55], with a large insertion of about 6 kb found within the coding sequence of these allelic variants [56]. Another blackleg R gene, Rlm13, was found to be located in a genomic region that has high numbers of structural variants such as SNPs, indels and PAVs on chromosome C03 [57]. On the other hand, a 700 bp deletion on B. napus chromosome C05 within a major QTL for V. longisporum resistance was identified through long-read sequencing [58], with gene PAV also found to be involved based on Illumina resequencing and Brassica 60K SNP analyses [59]. All of these evidences show that R genes in Brassica species evolve through highly dynamic processes of gene expansion and loss resulting in structural variations (SVs).

2.5. Complex Host-Pathogen Interaction

The level of a host’s susceptibility is often affected by its interaction with the pathogen, and the disease outcome could potentially rely on a complex interplay between the developmental stage of the host (cotyledon vs. adult stage) and environmental conditions, as shown in the glasshouse study of the AvrLmS-Lep2 Brassica blackleg pathosystem [60]. It has been shown that climate changes, such as variation in the CO₂ concentration, temperature fluctuations and water availability, play a role in influencing disease outcomes in host plants [61]. Disease severity is also impacted by the type of disease resistance of the host, whether it is qualitative or quantitative; in Brassica blackleg disease [62,63,64,65,66,67,68], or in the qualitative type, it depends on the specific pathotype that infects the host. For example, in the B. napus–Pyrenopeziza brassicae (light leaf spot) interaction, the host resistance is specific to the isolated pathotypes [69]. In the blackleg and clubroot pathosystems, the resistance genes of the host exert direct selection pressure on L. maculans and P. brassicae, thus increasing the adaptability of the pathogens and causing higher severity of disease in the host due to resistance breakdown, particularly when cultivars of major R genes are being deployed on large scales in fields [70,71]. To minimise sudden losses of R gene efficacy leading to major yield losses from L. maculans infection, growing Brassica varieties with different R genes on a rotational basis has been shown to be an effective strategy as a means to select against the corresponding virulence allele [72]. The tight interaction in the host–pathogen relationship, with the influence of physiological and environmental conditions, forces the Avr genes to evolve rapidly, thus leading to susceptibility of the host.

2.6. Complex Signalling Network Influencing Plant Immunity

Plant growth and defence is controlled by complex signalling pathways involving interplays of hormones, for instance, auxin, abscisic acid, brassinosteroids, cytokinin, ethylene, gibberellin, jasmonate and salicylic acid [73]. The B. napus valine-glutamine (VQ) genes, implicated in B. napus growth and development, were found to enhance resistance towards blackleg at the adult stage by interacting with the transcription factor WRKY, invoking the SA and JA signalling pathways, especially at the necrotrophic stage of L. maculans infection [74]. Abiotic stress hormones may also affect the biotic response pathways. In a study exploring the relationship between long-term (seven years) exposure to drought stress in B. rapa and the level of susceptibility of the plant towards Alternaria brassicae (Alternaria blackspot), a positive correlation of both variables was found with the involvement of the JA signalling pathway [75]. These examples demonstrate a complex signalling network in Brassica that influences R gene evolution and the extent of plant resistance developed over time. By taking advantage of some common genes along the signalling pathways, such as WRKY transcription factors, which have been found to play a role in many plant physiological systems—e.g., vernalisation in Chinese cabbage, or B. rapa [76], and Sclerotinia stem rot resistance in B. napus [77]—we increase the pool of candidate genes for gene pyramiding and stacking strategy, thus achieving the breeding of durable resistance in Brassica crops.

2.7. Epigenetics and R Gene Evolution

Since the beginning of this century, there has been an awareness that epigenetics can revolutionise medicine and agriculture [78]. Widely studied epigenetic modifications, such as DNA (de)methylation, histone post-translational acetylation, methylation or ubiquitination, chromatin assembly and RNA methylation, can regulate genomic activities such as chromatin density and gene expression/silencing, thus controlling varied phenotypes [79,80]. Experiments have revealed that epigenetic modifications play a role in transcriptional regulation of plant immunity against pathogens. Pathogen-infection-induced DNA hypomethylation results in elevated pathogen resistance, which is manifested by the interactions between Arabidopsis and bacteria, soybean and nematode, tobacco and virus, and Aegilops auschii and fungus [81,82,83,84]. In Arabidopsis resistance against the bacteria Pst DC300 and fungus Verticillium dahlia, histone H2B mono-ubiquitination (H2Bub1) plays a positive role [85,86]. Moreover, during ubiquitination in Arabidopsis, two novel ubiquitin E3 ligases, SNIPER1 and its homolog SNIPER2, were found to globally control the protein levels of sensor NLRs (sNLRs) reversely to maintain homeostasis and immune output [87]. All of these epigenetic modifications are important sources of genome evolution, participating in eukaryote genome regulation.

Among the studied epigenetic modifications, RNA N⁶-methyladosine (m⁶A) modification is the subject of the most up-to-date research, and it is the most prevalent internal post-transcriptional modification of mRNA. In experiments looking into the plant–pathogen arms race, results suggested that m⁶A modification is activated in pathogen-infected plants, which leads to different effects on the m⁶A and mRNA levels of genes related to plant–pathogen interaction, indicating a variety-specific m⁶A modification [88,89]. m⁶A has been shown to evolve synchronously with genome evolution and mRNA abundance [90]. According to Miao’s latest research on evolutionary analysis of m⁶A methylomes of 13 plant species representing evolution spanning over half a billion years, including A. thaliana in the family Brassicaceae, for plant R gene families, the m⁶A methylation ratio is negatively correlated with the number of family members. Furthermore, for all genes studied in the research and previous studies, the m⁶A methylation ratio is negatively correlated to genome size and gene members. Thus, the more expansion a genome experienced during its evolution, such as local gene duplication, the more m⁶A elimination it might induce. In addition, the earlier the orthologous genes split, the less diverse the m⁶A modification presents. Thus, the abundance of m⁶A can be used as a reference to determine the chronological order of gene evolution and isolation. This finding exposes new perspectives in the analysis of plant R gene evolution [91]. Although limited epigenetic studies of Brassica have been published, the intriguing correlation between m⁶A level and R gene evolution among Brassica and wild relatives suggests that R genes could be introduced into Brassica from such relatives with novel evolution.

2.8. Recessive Resistance Genes

In addition to the dominant R genes, recessive resistance genes also play an important role in plant host resistance [92]. Currently, there are two main hypotheses for the mechanism of these genes: according the first hypothesis, the dominant allele of a recessive resistance gene (also known as a susceptibility gene) might encode a specific host factor that is essential for a pathogen to complete its life cycle in plants. If the plant has the recessive resistance gene, it will lack or present a mutated version of the host factor, which makes the plant resistant to the pathogen [93]. The second hypothesis proposes that the recessive resistance gene might encode an inhibitor which interferes with some stage of the infection cycle [93].

Recessive resistance accounts for 50% of the 200 virus-resistance genes in crops [94], and to date, all the studied virus recessive resistances are governed by the abovementioned first hypothesis. The second hypothesis, on the other hand, can explain fungal recessive resistance [95]. Recessive resistance traits can be introduced into crop species by crossing, random mutagenesis, selection and genome editing [96,97]. It is proposed to be more durable than dominant resistance. The eukaryotic translation initiation factors (eIF) 4E and eIF4G and their isoforms (hereafter eIF4Es) are the most common recessive resistance genes identified to date. They are essential protein complexes involved in the translation of mRNA into proteins. They have been found in a range of plants to confer resistance to viruses, employing the first resistance hypothesis mentioned above: loss of susceptibility due to a deficiency of the eIF4Es gene. For example: eIF4Es-mediated resistance against viruses has been identified in A. thaliana [98,99,100], wild tomato (Solanum habrochaites) [101], lettuce (Lactuca sativa) [102], melon (Cucumis melo) [103], barley [104,105] and rice (Oryza sativa) [106], as well as in Brassica. In Brassica rapa, the recessive resistance gene retr01/retr02 was identified, which is an eIF4E-encoding gene associated with broad-spectrum resistance to turnip mosaic virus (TuMV) [107,108,109]. The resistance occurs due to a mis-splicing of the eIF4E allele [110].

In addition, another recessive resistance gene, retr03, encoding eIF2B was cloned in B. rapa [111]. In 2020, Shopan et al. reported that a total of 190 eIFs were detected in the B. juncea genome, 99 and 91 from the A and B subgenomes, respectively [112]. They were further clustered phylogenetically into 40 distinct subfamilies. Gene duplication plays an important role in the evolution and expansion of this eIF gene family. They identified a total of 33 duplicate gene pairs in the A subgenome and 35 pairs of duplicate/triplicate genes in the B subgenome of B. juncea [112]. After a duplication event, some gene copies were retained, owing to their critical function, while some genes were lost due to functional redundancy. The eIFs are highly conserved in Brassicaceae, with nearly 60% identity in A. thalian and B. juncea orthologs [112]. The orthologs did not diverge significantly between the A and B subgenomes of B. juncea [112].

MLO (mildew resistance locus O) is another class of recessive resistance gene. It also employs the first mechanism by encoding a negative regulator of plant immunity to achieve infection. Conversely, its mutation leads to broad-spectrum, high-efficiency and lasting resistance to disease in plants. It was first discovered in barley (Hordeum vulgare L.) with resistance to powdery mildew [113]. In B. nigra, a recessive resistance gene lm1 conferring resistance to blackleg (L. maculans) was classified as an MLO gene [114]. Yan et al. [115] explored the MLO gene families in A. thaliana, B. rapa, B. oleracea and B. napus. A total of 123 MLO genes were identified, which included 15 in A. thaliana, 23 in B. rapa, 28 in B. oleracea and 57 in B. napus [115]. Evolutionary analysis found that these 123 MLO genes were clustered into three different subgroups. Through comparative genome analysis, they found only 2 out of 15 A. thaliana MLO genes—MLO3 and MLO9—had no homologous genes in Brassica. Further comparison of MLO genes in A. thaliana and Brassica species identified that most of the MLO genes in A. thaliana had expanded in Brassica species, except MLO3 and MLO9. Among the transmembrane motifs of the 123 MLO genes, 71 genes have more than seven common transmembrane motifs. Conserved domains of these 71 MLO genes had one conserved amino acid sequence (462 aa long), which might be the main functional domain [115]. In addition to dominant resistance, recessive resistance genes also account for a great amount of disease resistance in crops. In Brassica, they mainly evolve through gene duplication.

The mechanisms governing R gene evolution in Brassica as discussed in this section have shed light on managing various diseases to increase crop production (

Table 1. Summary of the eight mechanisms of R gene evolution and their impact on disease management and crop production.

Mechanism of R Gene Evolution	Main Findings	Impact on Disease Management and Crop Production
1. Polyploid ancestry	Wild, weedy and domesticated types of Brassica species proposed and identified Comparative genomics of Brassica genomes revealed the genome relatedness between each species and their close relatives	Novel R genes sources can be identified from different varieties of Brassica species, especially the wild type Acceleration of interspecific hybridisation process in disease-resistant Brassica breeding
2. Disease resistance genes from introgression lines	Genome compatibility is established, and genome patterns after hybridisation can be studied using a pangenome approach	Produce varieties with hybrid vigour that can better resist various Brassica pathogens
3. Studying disease resistance genes from close relatives of Brassica	R gene homologs explored Immune response in Arabidopsis and other close relatives observed and compared with Brassica	Increase the gene pool of R genes for breeding purposes Speculate host–pathogen interaction and immunity response in different pathosystems
4. Structural variation of Brassica resistance genes	R gene behaviour studied, e.g., tandem and segmental duplications, allelic variants, SNPs, Indels, PAVs, etc.	Facilitate cloning of R genes using pangenome approach
5. Complex host–pathogen interaction	Factors affecting disease severity determined Examples of several pathosystems showed tight “arms- race” interaction influencing R gene evolution (Section 3)	Assist in pathotype and host resistance screening Develop better strategy for quality resistance crop breeding
6. Complex signalling network influencing plant immunity	Other genes such as those involving in growth and development are associated with immunity.	Enhance potential gene pyramiding strategy for more durable resistance
7. Epigenetics and R gene evolution	The m6A methylation ratio is negatively correlated with the number of R gene family members. The abundance of m6A can be used as a reference to determine the chronological order of R gene evolution and isolation.	Expose new perspectives in the analysis of plant R gene evolution
8. Recessive resistance genes	The mutation of recessive resistance genes plays important role in plant immunity. Recessive resistance genes might evolve through gene duplication.	Provide sustainable wide-spectrum resistance in crops

).

3. R Gene Evolution in the Blackleg B. napus–L. maculans Pathosystem and the Impact on Disease Management

For blackleg resistance, 19 R genes/alleles have been identified in Brassica species, and 5 of them have been cloned [116]. In the tight arms race of the Brassica–L. maculans pathosystem, the host R genes are often located in recombination hotspots within the genome and within regions of complex structural variation to allow more flexibility of genetic changes in response to the selection of highly evolved virulent pathogen genotypes [117,118]. When the R gene fails to evolve in advance of the virulent pathogen, the efficacy of the R gene fails [119,120] as soon as within three years [121,122]. More often than not, the resistance breakdown is not necessarily due to the direct effect of R–Avr interaction but due to dual specificity of the avirulence gene [116], showing that close monitoring of L. maculans pathotype screening and disease incidence in the field are equally important in managing blackleg disease.

Identification of Brassica blackleg R genes from different varieties such as breeding lines, cultivars, wild relatives, landraces, feral species and possibly ancient forms obtained from germplasm sources could enrich the gene pool of R genes. With advanced biotechnologies and the integration of genomics, phenomics and machine learning [123], supported by revolutionary CRISPR/Cas9 and other modern cloning and genetic transformation techniques [124,125,126], breeding disease-resistant Brassica varieties at accelerated speed is no longer a major constraint. The main challenge, however, is the implementation of the R genes in each cultivar and the deployment strategy of these cultivars in the field, taking into consideration the evolutionary mechanisms between the two players of the blackleg pathosystem, to ensure durability of R genes.

It was recently shown through a simulation study that rotating B. napus cultivars containing a single R gene with different R genes every five years and pyramiding two genes in a cultivar, with two different pyramided genes changing every year, in two-year rotations are effective strategies to extend the durability of R genes [72]. This indicates that the greater the R gene resources, the higher the chance of successful implementation of crop rotation and stacking. In addition, the B. napus resistance outcome is dependent on the genetic background of the host, as it was seen that within the same R gene–L. maculans effector interaction in different genotypes (R gene introgressed lines), the dynamics of the gene expression of defence-related genes was different; hence, the intensity of immunity was also different [127]. This has practical implications for Brassica disease management, where these defence-related genes that are differentially expressed in genotypes with different genetic background can be used as markers to determine which parental genotype will be most suitable when breeding R-gene-resistant B. napus. Another effective strategy to deploy R genes in the field is to breed B. napus cultivars that have qualitative R gene resistance in combination with quantitative resistance [128]. The biology behind this mixture of resistance types is that there would be higher genetic variation in the host population, thus disrupting the selection pressure on L. maculans and decreasing the evolution frequency of virulent pathotypes [117,129]. Some other ways to control the pathogen population include agricultural practices such as applying fungicide, crop isolation and treating the infected cultivar residue by burning, tillage or burial [130].

4. Technologies to Study Evolutionary Origins of Brassica

4.1. Genome Level

Structural variation (SV) plays an important role in the evolution of plant genomes. Large-scale structural variants can be detected using advanced third-generation sequencing methods and read mapping strategies such as high throughput chromosome conformation capture (Hi-C) and BioNano optical mapping technologies, as well as third-generation whole-genome sequencing strategies such as PacBio single-molecule real-time sequencing (SMRT) and Oxford Nanopore (ONT), which produce long reads and have been shown to overcome limitations on short reads to detect various SVs [131,132]. For example, in B. oleracea, large (100 kb or greater) SVs were detected using a combination of PacBio SMRT (DNA sequencing), PacBio Iso-Seq (RNA sequencing), BioNano optical mapping and Hi-C technologies [133], with genes related to stress response found to be related to these SVs. In B. napus, a 700 bp deletion of the V. longisporum resistance gene was identified using PacBio ONT sequencing [58].

Whole-genome sequencing, RNA sequencing, flow cytometry and some cytogenetic technologies such as fluorescence in situ hybridisation/genomic in situ hybridisation (FISH/GISH), depending on hybridisation between fluorescent probe DNA and introduced target DNA, help to reveal the process of allopolyploid formation and alien chromosome fragment introgression, like the artificial creation of hexaploid/octoploid Brassica and natural development of other allopolyploid species like tetraploid cotton [134,135,136]. Using advanced sequencing and mapping tools on the complex polyploid Brassica genomes formed through interspecific hybridisation, either through artificial or spontaneous means, will allow us to understand plant genome evolution at greater depth and identify clearly the valuable genes underlying the genomes in each of the Brassica species [137].

At the gene level, candidate gene identification through bulked segregant analysis (BSA), whole-genome resequencing (WGRS) and QTL-seq methods have been shown to be very effective, delimiting the QTL interval of the candidate region in Brassica plants—for example, in flower colour candidate gene determination in B. juncea [138], as well as flowering and trichome formation and shoot branching in B. rapa [139,140]. The BSA-RNA sequencing approach was undertaken to identify resistance genes for clubroot P. brassicae in B. oleracea [141]. Some of the main advantages of using the BSA-QTL-seq strategy for candidate gene identification include being highly robust, as it can be applied not only on Brassica plants but fungal isolates such as L. maculans, and only a limited number of representatives (10 samples) from the segregating population are required instead of hundreds of individuals for genotyping and phenotyping work [60]. The Brassica 60K Illumina Infinium array and WGRS approaches using high-quality Brassica genome assemblies are useful tools to explore SNPs at the whole-genome level. For example, WGRS on 991 B. napus accessions representing various ecotypes identified more than 50 candidate NLR genes located within highly evolved genomic regions of B. napus [44]. A more targeted approach towards candidate R gene identification using the “bait capturing” method to capture all RGA-like sequences within the genome, called RenSeq (R gene enrichment sequencing) and RLP/KSeq (receptor-like protein/kinase enrichment sequencing) [142,143], collectively referred to as RGASeq, can be useful in large, complex genomes like Brassica [42]. All of these advanced NGS tools offer high-resolution mapping and quick discovery of novel R genes for the study of the evolutionary history of R genes for resistance breeding.

4.2. Pangenome Level

With more sequenced Brassica genomes available, the pangenome approach has become more attractive, considering the rich evolutionary information that can be deciphered at the genome and gene levels. It is currently feasible to use pangenomes with the implementation of state-of-the-art machine learning, not only to discover all the genes present in different Brassica species and explore the genome changes that happen in each of these polyploid species during the evolutionary pathway, but also to use this wealth of information in genomic selection, which offers a very promising future for Brassica crop breeding [144]. By developing pangenomes of B. napus and its progenitors, B. oleracea and B. rapa, it was revealed that defence- and stress-related genes are common dispensable genes and that gene loss events observed in specific Brassica plants are attributed to various evolutionary mechanisms [48]. A B. napus pangenome database was recently built, called BnPIR, which is a one-stop platform for users to look for pangenome resources using a built-in browser, making possible the comprehensive study of evolution-related variations in B. napus [145].

4.3. Epigenetic Level

There has been rapid development in high-throughput epigenetic and epitranscriptomic sequencing in recent years. Whole-genome bisulfite sequencing (WGBS), targeted methylome sequencing (TMS), such as shotgun bisulfite sequencing, and post-bisulfite adaptor tagging (PBAT)-assisted TMS are used in DNA methylome analysis [146,147]. Enriching abundant m⁶A-modified RNA fragments through immunoprecipitation with m⁶A-specific antibodies and technology like methylated RNA immunoprecipitation with next-generation sequencing (MeRIP-seq) has been widely applied to enable transcriptome-wide profiling of RNA m⁶A modification related to pathogen resistance or genome evolution [91,148]. The data analysis is carried out using bioinformatic technologies to reveal the epigenetic and epitranscriptomic changes formed during adaption to environmental changes like biotic and abiotic stress.

4.4. High-Throughput Phenotyping

Association analysis between plant genotypes and phenotypes is an important method to understand the differences in crucial traits regulated by genetic variations formed during plant genome evolution and development. As we reviewed above, there has already been major development in technologies to acquire abundant genetic and genomic data. However, the traditional labour- and time-consuming phenotyping methods remain a bottleneck in association analysis to understand plant genome development [149].

Nowadays, there are some high-throughput phenotyping technologies that have been introduced into plant improvement which allow data acquisition in a rapid and non-invasive way. These high-throughput phenotyping methods always rely on dynamic optical imaging equipment and machine learning [150]. Take its application in pathogen resistance, for instance: hyperspectral imaging was used to assess disease severity, such as cellular level changes in barley leaves infected with powdery mildew [151,152]; an automated imaging scanner was used to assess resistance against Septoria tritici blotch (STB, caused by fungus Zymoseptoria tritici) in wheat in a field experiment, with 26 chromosomal intervals harbouring several novel loci identified, showing quantitative resistance emerging from co-evolution with the pathogen [153]. High-throughput digital imaging and R scripts were conducted in the genome-wide association study (GWAS) of Arabidopsis against Botrytis cinerea, identifying 23 candidate genes [154]. All of these high-throughput phenotyping technologies present potential to help us further understand the coevolution between plants and pathogens.

5. Conclusions

The Brassica R genes have evolved through many evolutionary forces. With such a deep history of hybridisation and polyploidy formation in Brassica species, we believe many more novel evolutionary forces and novel R genes have yet to be uncovered (Figure 2). With currently available high-quality Brassica genome assemblies and pangenome resources and many more adaptations of NGS in DNA, RNA or methylation sequencing using sophisticated bioinformatics algorithm, coupled with the application of advanced flow cytometry and cytogenetical technologies such as FISH and GISH, R gene evolution and discovery in Brassica will be accelerated in the near future. High-throughput phenotyping approaches exploited across all Brassica crops, along with continuous efforts to characterise pathogen isolates, will further enhance our knowledge of host–pathogen interaction for breeding high-quality Brassica crops.

Furthermore, the newly confirmed synergy of PTI and ETI in Arabidopsis might show a novel point in the study of plant R gene evolution. The heredity and epigenetic memory might show new perspectives in crop resistance breeding, with a possibility that the m⁶A modification ratio reflects R gene evolution conservation. Understanding R gene evolution and the genetic diversity across crops’ wild relatives may also help in the introgression of novel R genes for sustainable breeding of high-quality Brassica crops.

Compared with major R genes, which evolve quickly to win the race against pathogens, the resistance guaranteed by recessive resistance genes is rather conserved. This conservation provides a sustainable, wide-spectrum resistance. The mutation of recessive R genes can be considered as a method in engineering broad-spectrum resistance in crops, together with the utilisation of E3 ligases controlling global sNLRs in plants.

There are still many pieces of puzzles that need to be researched and understood to achieve a complete picture of the plant immune system and its evolution. With the development of high-throughput genotyping and other advanced gene validation tools such as CRISPR/Cas9, achieving a near-complete understanding of R gene evolution and precision breeding is very promising. In this manner, it is favourable to achieve higher accuracy in the study of plant R gene evolution and of breeding for broad-spectrum resistance in crops through advanced methods, causing fewer public concerns.