Host–parasite interactions in clubroot of crucifers

Abstract

Clubroot, caused by Plasmodiophora brassicae, is an important root disease of crucifers worldwide. In this review, the molecular aspects of clubroot pathogenesis and resistance are discussed. Topics covered include recent studies on the processes associated with infection by primary and secondary pathogen zoospores, examination of the expression patterns of P. brassicae genes at different stages of infection, and the concurrent identification of candidate genes for functional studies. Although the whole genome sequence of P. brassicae is not yet available, molecular studies of the pathogen are nonetheless moving into the genomics era. Transcriptomics and functional analyses of individual genes involved in the P. brassicae/Brassica interaction have recently been conducted to complement studies on the excellent model host, Arabidopsis thaliana. Two resistance genes have been cloned and their molecular functions illustrated, supporting suggestions that the clubroot pathosystem follows the gene-for-gene model. Although much about the pathogenesis of clubroot remains unknown, significant advances have been made in recent years, which may facilitate the identification of targets for agro-chemical development and resistance breeding.

Résumé

La hernie, causée par Plasmodiophora brassicae, est une importante maladie de la racine chez les crucifères, et ce, à l’échelle mondiale. Dans cette revue, nous discutons des aspects moléculaires de la pathogenèse et de la résistance de la hernie. Les sujets traités incluent de récentes études effectuées sur les processus associés aux infections primaires et secondaires causées par des zoospores pathogènes; l’examen de l’expression génétique de P. brassicae à différents stades de l’infection; et l’identification concurrente de gènes candidats utilisés dans les études fonctionnelles. Bien que nous ne disposions pas encore de la séquence complète du génome de P. brassicae, les études moléculaires menées sur l’agent pathogène se rapprochent néanmoins de l’ère de la génomique. La transcriptomique et des analyses fonctionnelles de gènes individuels impliqués dans l’interaction P. brassicae/Brassica ont récemment été utilisées pour compléter les études sur l’excellent hôte type qu’est Arabidopsis thaliana. Deux gènes de résistance ont été clonés et leurs fonctions moléculaires, illustrées, appuyant la suggestion stipulant que le pathosystème de la hernie est conforme au modèle « gène pour gène ». Bien que nous ne connaissions pas entièrement la pathogenèse de la hernie, d’importantes avancées récentes faciliteront peut-être l’identification de cibles pour le développement agrochimique et la sélection en vue de la résistance.

Keywords:

• canola
• pathogenesis
• pathogenicity
• Plasmodiophora brassicae
• resistance

Mots clés:

• Canola
• pathogenèse
• pathogénicité
• Plasmodiophora brassicae
• résistance

Introduction

Plasmodiophora brassicae Woronin causes clubroot disease in cruciferous plants, and is an emerging threat to Canadian canola (Brassica napus L.) production (Hwang et al. 2012). Among the pathogens causing major plant diseases, P. brassicae has a distinct intracellular lifestyle. This intracellular lifestyle, coupled with the agricultural importance of clubroot worldwide, has warranted a thorough study of P. brassicae by a number of research groups. However, the obligate biotrophic nature of P. brassicae and the difficulty associated with obtaining purified pathogen isolates have prevented the application of many techniques used in other pathosystems. The infection process of P. brassicae is not clearly understood and ‘-omics’ level research is only now being initiated. Fewer than 200 non-redundant expressed sequence tags and protein sequences are available in the database and only a few genes have been investigated on a transcription or protein level. From an applied perspective, a few resistance genes have been identified in cruciferous hosts (Piao et al. 2009). Among them, two were cloned and their molecular function studied (Ueno et al. 2012; Hatakeyama et al. 2013). These new data on the molecular basis of resistance, together with data from studies on pathogenicity, suggest that the interaction of P. brassicae and its hosts may be compatible with the gene-for-gene theory, in which pattern recognition receptors (PRRs), nucleotide-binding and leucine-rich repeat (NB-LRR) proteins and effectors, which are all core components operating in host immunity in other plant pathosystems (Jones & Dangl 2006), determine the outcomes of the interplay between P. brassicae and its hosts.

Most previous work on the molecular basis of clubroot disease was conducted on Arabidopsis thaliana (L.) Heynh. or cruciferous vegetables. In the past several years, the outbreak of clubroot on canola (B. napus) in western Canada has encouraged significant interest in the interactions of P. brassicae with this economically important host. Thus, this review focuses on research advances with respect to understanding the interaction between P. brassicae and its canola host. For researchers new to P. brassicae, comprehensive reviews of the molecular basis of the clubroot pathosystem can be found in Ludwig-Müller et al. (2009), Piao et al. (2009), Siemens et al. (2009a) and Hwang et al. (2012).

Part I: Clubroot pathogenesis

Evolutionary aspects of P. brassicae and clubroot host–parasite interactions

Despite intensive searches in the Mediterranean centre of diversity of several brassica hosts, the corresponding centre of diversity of P. brassicae has been difficult to locate (Dixon 2009a). Seemingly, clubroot disease is associated almost exclusively with agriculture, suggesting that it may be a ‘disease of cultivation’. As an obligate biotroph, P. brassicae should have been in a continuous arms race with its hosts in the agroecosystem and this would involve repeated fixation of alleles in both partners (Brown & Tellier 2011). This hypothesis provides a theoretical basis for the existence of a gene-for-gene model in the clubroot pathosystem, which is also supported by the accumulated experimental data. Firstly, there is physiological specialization (Honig 1931) and a clear pathotype structure in P. brassicae (Howard et al. 2010). Multiple pathotypes may be present in the same populations or even in the same root galls. Secondly, most identified clubroot resistance genes have been found to be race or pathotype-specific (Piao et al. 2009) and both of the two molecularly characterized resistance genes encode typical race-specific resistance proteins (Ueno et al. 2012; Hatakeyama et al. 2013).

Current status of P. brassicae pathogenicity studies

For any particular plant pathosystem, there are three eras in molecular plant pathology research: (1) the ‘grind and find’ era, in which virulence molecules are isolated from cell-free extracts, (2) the ‘screen for gene’ era, which is also known as the molecular genetics era and is focused on one or a few genes, usually studied by marker-insertion mutagenesis or targeted gene knockout, and (3) the ‘patterns that matter’ era, in which genomics technique are utilized to globally identify pathogen-related genes (Schneider & Collmer 2010). Based on the current literature, molecular studies on clubroot pathogenesis appear to be in the second era, albeit the biotrophic nature of P. brassicae has hampered the utilization of most molecular tools. The third stage is beginning and it is anticipated that the whole genome sequence of P. brassicae will soon be available (Dixelius et al. Swedish University of Agricultural Sciences, and Borhan et al. Agriculture and Agri-Food Canada, personal communication).

Pathogenicity-related genes and proteins identified from P. brassicae

To date, only a limited number of partial cDNA fragments or full-length clones have been isolated from P. brassicae. A search of the National Center for Biotechnology Information (NCBI) database (24 October 2013) for P. brassicae-originated accessions yielded 180 nucleotide sequences, 138 expressed sequence tags and 53 proteins. Of these sequences, only a small number have a potential importance in pathogenicity, largely based on sequence similarity with genes from other species. Most of these sequences were derived from cDNA libraries constructed by suppression subtractive hybridization (SSH). Among them are 76 P. brassicae genes isolated from a SSH library constructed between RNA from P. brassicae-infected and uninfected Arabidopsis tissue (Bulman et al. 2006). Many of the sequences were predicted to contain signal peptides for translocation of the encoded proteins. Using Brassica rapa L. clubroot galls, Sundelin et al. (2011) isolated about 140 clones. Half of these clones, which represented 24 unisequences (10 of which were newly characterized), originated from P. brassicae. Through the use of next generation sequencing (NGS) technologies, Burki et al. (2010) generated 3167 P. brassicae specific contigs from infected B. rapa. These sequences, however, were very short and their functions or relevance to clubroot development was not described. Feng et al. (2010) constructed a SSH library at the early stages of infection (7 days after inoculation) of canola roots by P. brassicae. The library consisted of 797 clones that represented 439 unigenes. Thirty-two of these genes were demonstrated to be of a P. brassicae origin, and of these 24 had not been previously reported. Differential expression of a subset of these genes in the resting spores and at 7 days after inoculation was observed by quantitative PCR (qPCR), suggesting that some of these genes may play a role during pathogenesis.

In addition to these libraries, a few other genes have been identified from P. brassicae and postulated to be related to its pathogenicity. Ito et al. (1997) identified a gene (Y10) whose expression was exclusively correlated with the vegetative plasmodial stage of P. brassicae. Brodmann et al. (2002) found a trehalose-6-phosphate synthase gene with expression correlated with an accumulation of trehalose in resting spores. Ando et al. (2006) demonstrated that the expression of PbSTKL1 increased strongly beginning 30 days after inoculation and was coincident with resting spore formation. Two cDNAs, PbBrip9 and PbCC249, were found to be strongly expressed at stages of disease development corresponding to the occurrence of sporulating plasmodia (Siemens et al. 2009b). Although no data are available regarding their biological functions, all these genes may have a potential role in clubroot pathogenesis.

Until now, only one P. brassicae gene, a serine protease (PRO1), has been functionally studied and proven experimentally to be important for pathogenesis (Feng et al. 2010). Southern analysis and specific PCR amplification indicated that PRO1 is a single-copy gene present in a broad range of P. brassicae pathotypes. Northern analysis revealed expression of PRO1 during plant infection, as well as in dormant and germinating resting spores. The expression pattern assessed by Northern analysis was further confirmed by qPCR in a later study (Feng et al. 2013a). The open reading frame of PRO1 was transferred into Pichia pastoris (Guillierm.) Phaff and Pro1 proteins were heterologously produced. Pro1 showed proteolytic activity on skimmed milk and Suc-AAF-AMC, and the activity could be inhibited by serine protease inhibitors and the chelating agent EDTA. The optimal temperature of Pro1 was 25 °C, and it exhibited maximum activity at pH 6.2–6.4 and minimal activity at pH 6.8–7.6. These values coincide with the temperature and pH conditions favourable for P. brassicae resting spore germination in the field. When Pro1 was used to treat canola root exudates, it enhanced the stimulatory effect of the root exudates on P. brassicae resting spore germination, indicating that Pro1 plays a role during clubroot pathogenesis by stimulating resting spore germination through its proteolytic activity.

Preliminary genomic studies on P. brassicae

A few studies have been conducted on P. brassicae that generated data for what might be termed genomics. Bulman et al. (2007) identified DNA sequences spanning or flanking 24 P. brassicae genes, eventually sequencing close to 44 kb of genomic DNA. Some general characters of P. brassicae genes emerged from this preliminary genome survey: (1) P. brassicae genes are rich in spliceosomal introns with consensus splice sites and branch-point sequences similar to those found in other eukaryotes, (2) as a consequence of being intron-rich, splicing is an important feature of gene expression, (3) P. brassicae transcription is likely to begin from initiator elements rather than TATA-box containing promoters, and (4) intergenic distances are short, ranging from 44 to 470 bp.

Recently, Feng et al. (2013a) reported a gene expression study in which all database available P. brassicae genes (118) were investigated by dot-blot hybridization and qPCR. The expression of these genes in primary and secondary zoospores was compared. Both dot blot and qPCR identified up- and down-regulated genes and the correlation between these two techniques was confirmed. Quantitative PCR indicated that 58 genes were up-regulated in the secondary zoospores relative to the primary zoospores, whereas 55 were down-regulated.

Tools for molecular studies of P. brassicae

Resting spore purification and single spore isolation

Currently, there is not a standard resting spore purification protocol that satisfies the requirement for molecular studies. The routinely used protocols as described by, for example, Bryngelsson et al. (1988) and Asano et al. (1999), are sufficient in quality for inoculation purposes, but not for DNA or RNA extraction when contamination from other organisms needs to be excluded. Similarly, although there are several reports on the isolation of single-spores of P. brassicae (Xue et al. 2008 and references therein), the techniques described are not capable of producing sufficient volumes of material. Therefore, while these techniques may be suitable for many purposes, further improvements are needed to develop protocols that would enable the purification and recovery of much larger numbers of single resting spores, which would increase efficiency and facilitate molecular studies.

Isolation of zoospores

A protocol to isolate secondary zoospores was reported by Feng et al. (2012b). By washing the roots of infected canola seedlings 7 days after inoculation, the authors were able to isolate secondary zoospores without contamination of resting spores and demonstrated that the collected secondary zoospores could initiate infection. This technique allowed investigation of the infection processes of primary and secondary zoospores separately, and will be useful in genetic studies in which resistance can be partitioned into different types based on the infection stage at which it is expressed. This study has stimulated several additional studies to investigate P. brassicae gene expression (Feng et al. 2013a) and the plant responses to primary and secondary infection (Deora et al. 2012; Feng et al. 2013c).

In vitro dual culture. Bulman et al. (2011) standardized an in vitro dual culture system to study the genomics of P. brassicae and related protist species. In the absence of exogenous plant growth regulators, stable, long-term B. rapa cell callus cultures infected by P. brassicae were produced and demonstrated to be an excellent starting material for gene discovery. By using this technique on a related plasmodiophorid species, Spongospora subterranea (Wallr.) Lagerh., the authors constructed a pilot-scale DNA library and found that almost all of the DNA clones were from the protist rather than the plant host. This technique provides an avenue to remove plant contaminants from the pathogen materials, which is critical in genomics studies.

Genetic transformation

A protocol for the genetic transformation of P. brassicae has been recently reported (Feng et al. 2013b), in which a protoplast preparation was superseded with lithium acetate treatment and the selection step was omitted. With this protocol, the authors transformed germinating/germinated resting spores with two fungal expression vectors and then inoculated them on canola plants. Approximately 50% of the galls obtained contained resting spores from which transforming DNA could be detected by PCR analysis. Quantitative PCR and genome walking indicated that the transforming DNA was integrated into the P. brassicae genome. Transcript of the transforming gene was also detected by reverse transcription qPCR from selected transformants. Following another round of inoculation, the transforming DNA could still be identified from the resultant galls, indicating the stable inheritance of the transforming DNA. This protocol makes strain labelling and gene knockout studies possible in P. brassicae.

Studies on P. brassicae

Non-host infection

The behaviour of the pathogen on a non-host is a desirable research area because it is directly related to non-host resistance in the plant. Primary infection by P. brassicae has been observed in non-host plants by a few authors, although secondary infection is not regularly observed (MacFarlane 1952) and when observed, no resting spores are produced (Ludwig-Müller et al. 1999).

Feng et al. (2012b) inoculated secondary zoospores of P. brassicae collected from infected root hairs of the host canola and non-host ryegrass (Lolium perenne L.) onto healthy roots of both plant species. The experiment consisted of all possible combinations of the two plant species and the two sources of inoculum. Primary infection of the root hairs and secondary infection of the cortex were similar in all of the treatments. The degree of secondary infection and extent of secondary plasmodia development were higher in canola inoculated with zoospores from canola than in ryegrass inoculated with zoospores from ryegrass. At 35 days after inoculation, typical clubs developed on the canola plants inoculated with secondary zoospores from canola, and tiny clubs developed on the canola plants inoculated with zoospores from ryegrass. Secondary infection occurred in about one-third of the ryegrass plants but no clubs developed, regardless of inoculum source. A lack of gall development in ryegrass was not unexpected, since the fibrously rooted ryegrass lacks the lateral meristems involved in gall development (Rennie et al. 2013). This was the first demonstration that secondary zoospores produced on a non-host can infect a host and confirms that secondary infection can occur in a non-host. The data were consistent with the previous notion that host specificity may be less in the root hairs compared with the cortical cells (Kageyama & Asano 2009) and further suggested that on non-hosts, primary infection can induce plant resistance to secondary infection.

Host recognition

Resting spores can persist in the soil for as long as 20 years (Wallenhammar 1996). From an evolutionary point of view, the production of large numbers of persistent spores is a fitness trade-off for being soil-borne. It is reasonable to infer that environmental factors are critical to stimulating resting spore germination. These factors include not only the physical and chemical components of the soil environment, but also biological signals from the host (Dixon 2009b). Thus, host recognition by the pathogen will be important for clubroot pathogenesis and will be an interesting area for research.

Root exudates, from host as well as non-host plants, have been reported to trigger the germination of P. brassicae resting spores (Suzuki et al. 1992; Friberg et al. 2005; Niwa et al. 2008), and the term ‘germination stimulating factors’ (GSFs) has been assigned to these triggering factors (Suzuki et al. 1992). GSFs consist of two components: (1) compounds that can be produced by both host and non-host, and (2) compounds unique to the host. It is probable that caffeic acid, coumalic acid and corilagin, identified by Ohi et al. (2003), are among the first components. For the second GSF component, the involvement of pathogen enzymatic activity has been proposed. Feng et al. (2010) confirmed that spore germination was enhanced by host root exudates and showed that treatment of root exudates with the pathogen-derived protease Pro1 increased the effect on spore germination stimulation. Thus, it appears that the proteolytic products resulting from the Pro1 treatment enhance the spore germination stimulating function of the root exudates. These data indicate that Pro1 not only generates stimulatory chemicals, but also provides a potential avenue for the pathogen to recognize its host.

Primary and secondary infections

Partitioning of the primary and secondary infections is an appealing research area because different mechanisms may be involved in the infection process and in the host resistance response at each stage. Two recent studies indicate that the primary and secondary zoospores may be functionally similar (Feng et al. 2013c) and differ only in patterns of gene expression (Feng et al. 2013a). In the former study, direct inoculation onto plants of secondary zoospores resulted in primary infections similar to those obtained with resting spores. When the plants were inoculated with a high concentration (1 × 10⁷ mL⁻¹) of resting spores 2 days after being challenged with a low concentration (1 × 10⁴ mL⁻¹ or 1 × 10⁵ mL⁻¹) of resting spores, secondary infections were observed earlier than the secondary infections resulting from inoculation with the high concentration alone, and were more severe than those produced by inoculation with the low concentration alone. Compared with single inoculations, secondary infections on plants that had received both inoculations remained at higher levels throughout a 7-day time course. These data suggest that primary zoospores can directly cause secondary infection when the host has already undergone primary infection. Based on these observations, the authors argued that primary infection, rather than the primary zoospore, is a prerequisite for secondary infection. The interplay between P. brassicae and its host during primary infection may be critical for determining resistance or susceptibility. This argument is supported by a previous study in which a root hair defective mutant of Arabidopsis showed tolerance to clubroot (Siemens et al. 2002).

In Feng et al. (2013a), the gene expression patterns in primary and secondary zoospores were investigated by dot-blot and qPCR. The data indicated that the majority of the investigated genes were differentially expressed in primary and secondary zoospores. Among the 117 genes assessed, only four remained unchanged in their expression, as indicated by qPCR.

Part II: Host resistance

Transcriptome studies

Studies on pathogen-induced changes in host gene expression and metabolism have also been hampered by the difficulties associated with working with this obligate parasite. A few studies used transcriptomic and proteomic approaches to elucidate the mechanisms of P. brassicae infection and coordinate plant reactions. Using a full-genome Affymetrix chip, Siemens et al. (2006) found that more than 1000 Arabidopsis genes were differentially expressed in P. brassicae-infected roots at either 10 or 23 days after inoculation versus non-inoculated roots. These included genes associated with growth and cell cycle, sugar phosphate metabolism, and defence. The involvement of plant hormones in club development was further supported; genes involved in auxin homeostasis, such as nitrilases and members of the GH3 family, were up-regulated, whereas genes involved in cytokinin homeostasis were strongly down-regulated at both time points. Cytokinin oxidase/dehydrogenase overexpressing lines were disease resistant, clearly indicating the importance of cytokinin as a key factor in clubroot disease development. Another interesting finding from this study is that most known defence- or resistance-related genes were either not differentially expressed or down-regulated at 23 vs. 10 days after inoculation. This was supported by Feng et al. (2012a), in which expression of several canola genes homologous to Arabidopsis genes found to be down-regulated at 35 days after inoculation, were highly expressed at early stages (7 days after inoculation) and down-regulated at later stages (42 days after inoculation) of infection. At the later stages of infection, the galls were forming and the resting spores were mature. Thus, the host genes involved in resistance/susceptibility may no longer be as active as during the early infection stages. Another explanation is that during the later stages, the pathogen already has successfully suppressed the expression of most resistance-related genes.

Proteomic studies

Proteome-level analysis, a powerful tool for high-throughput global protein expression analysis using two-dimensional gel electrophoresis (2-DE) coupled with mass spectrometry (MS) and bioinformatics, has been extensively used in the identification of proteins that are differentially expressed in plants in response to various abiotic or biotic stresses including those caused by plant pathogens. In order to investigate whether the pattern of specific host proteins are changed during clubroot formation, Hansen et al. (1994) employed 2-DE to compare the root protein profiles of Brassica oleracea L. with B. oleracea infected with P. brassicae four weeks after inoculation. They found that several proteins present in non-infected roots were absent or strongly reduced in the infected roots. However, the identities of these proteins were not determined. In a proteomic study of a Chinese cabbage cultivar and two P. brassicae isolates, Ito et al. (1996) were able to resolve 133 differentially abundant protein spots, relative to non-infected controls, in infected roots 3 to 4 weeks after inoculation. The 133 protein spots were further classified into 6 groups, including spots that were enhanced in the susceptible response (59 spots), unique to the susceptible response (6 spots), repressed in the susceptible response (9 spots), enhanced in the resistant response (35 spots), unique to the resistant response (5 spots), and repressed in the resistant response (19 spots). N-terminal amino acid sequencing revealed that one protein (25 kDa, pI 7.0) that was enhanced in the susceptible response group showed high homology with pathogenesis-related protein group 5. These results suggest that the patterns of gene expression are different in the susceptible and the resistant responses. Ideally, it would be very useful if the identities of all of the 133 protein spots had been determined in the study.

In a differential protein analysis of infected versus non-infected roots and hypocotyls of Arabidopsis with 2-DE and MS, Devos et al. (2006) reported that 12% (46/390) of the visualized Arabidopsis proteins showed an altered abundance 4 days after inoculation with P. brassicae compared with the non-inoculated plants, including proteins involved in metabolism, cell defence, cell differentiation and detoxification. The metabolism related enzymes adenosine kinase 2 (ADK2) and fructose-bisphosphate aldolase (FBA) were found to be down-regulated at 4 days after inoculation. Down-regulation of ADK2 during the early stages of P. brassicae infection may keep the elevated levels of active cytokinins high, which is consistent with the increased endogenous isopentenyl adenine content. Down-regulation of FBA upon P. brassicae infection suggests a flow toward the production of glucose, which correlates with the fact that galls of P. brassicae act as a carbon sink that may be induced by the high cytokine level present at the infection site (Devos et al. 2006). Myrosinase, an enzyme that can catalyse indole-3-methyl glucosinolate degradation to form indole-3-acetic acid, was also found to be 6 times more abundant in infected Arabidopsis at 4 days after inoculation. During P. brassicae infection, a number of detoxification related proteins were found to be up-regulated, including catalase, glutathione S-transferase, thioredoxin and ferredoxin-nitrate reductase. Of the 46 differentially regulated proteins, 11 proteins including ADK2, FBA, peroxiredoxin, α-tubulin and heat shock protein 70 were down-regulated, and 35 were up-regulated including myrosinase, glutathione S-transferase, ferredoxin-nitrite reductase and pectin methylesterase (Devos et al. 2006).

More recently, changes in the root protein profile were examined by 2-DE at 12, 24, 48 and 72 hours after inoculation of a susceptible canola (B. napus) genotype with P. brassicae (Cao et al. 2008). A total of 20 protein spots were identified as either up-regulated (13 spots) or down-regulated (7 spots). Decreased abundance of adenosine kinase, which is involved in cytokine homeostasis, supported the previous reports that cytokinins play a key role in the early phases of P. brassicae infection (Devos et al. 2006). An approximately 6-fold reduction in caffeoyl-CoA O-methyltransferase abundance suggested a reduction in host lignin biosynthesis after inoculation, and is consistent with the compatible nature of the B. napus/P. brassicae interaction examined. Levels of enzymes involved in the metabolism of reactive oxygen species, such as copper/zinc superoxide dismutase and cytochrome c oxidase, declined sharply at 12 hours after inoculation, but increased at 24–72 hours. Protein identification by MS analysis revealed that the 20 differentially regulated proteins could be classified into proteins involved in lignin biosynthesis, cytokinin metabolism, glycolysis, intracellular calcium homeostasis, and the detoxification of reactive oxygen species.

These studies of the Brassica/P. brassicae interaction at the proteome level have provided researchers with new insights into clubroot pathogenesis and resistance. However, functional validation of the differentially abundant proteins is still needed, in order to elucidate more clearly their roles in the clubroot pathosystem.

Non-host and basal resistance

Non-host resistance is the resistance shown by an entire plant species to all genetic variants of a pathogen (Mysore & Ryu 2004). In this context, the pathogen is referred to as non-adapted to the plant. Basal resistance is the first line of defence of a host to the adapted pathogen, which is achieved through a set of defined receptors, also referred to as pattern recognition receptors (PRRs), that recognize conserved microbe-associated molecular patterns (MAMPs) (Jones & Dangl 2006; de Wit 2007). Upon MAMP recognition, primary defence responses are induced such as cell wall alterations, deposition of cellulose and the accumulation of defence-related proteins (van Loon et al. 2006). A new conceptual framework has been recently proposed that non-host resistance and basal resistance to adapted pathogens may rest on similar principles (Niks & Marcel 2009; Fan & Doerner 2012). Based on studies of other pathosystems, plant primary metabolism has been demonstrated to play a significant role in non-host and basal resistance (Stuttmann et al. 2011).

The involvement of plant primary metabolism in clubroot pathogenesis has been studied, with the induction of sink metabolism regarded as a non-specific response of host roots to P. brassicae infection (Ludwig-Müller et al. 2009). In P. brassicae-infected Arabidopsis roots, expression of sucrose synthase and starch synthase is induced (Siemens et al. 2006). The lower rate of photosynthesis is associated with an increase in invertase activity in the roots, and an accumulation of hexoses with a down-regulation of photosynthetic gene expression is observed (Siemens et al. 2011). It is reasonable to postulate that plant genotypes with down-regulated sucrose/starch synthase or invertase activity would carry a basal resistance to the pathogen.

This hypothesis was partially supported by Gravot et al. (2011). These authors proposed that the induction of plant trehalase during clubroot disease in Arabidopsis was a defence mechanism, which could help the plant cope with the accumulation of pathogen-synthesized trehalose. This mechanism of resistance is also found in the susceptible Arabidopsis accession Col-0, and the inhibition of trehalase activity during clubroot infection resulted in enhanced disease symptoms, suggesting that trehalase induction is a basal defence mechanism: a disease-resistance mechanism activated by a virulent pathogen on a susceptible host (Jones & Dangl 2006). Gravot et al. (2011) argued that pathogen-derived trehalose can be viewed as a microbe-associated molecular pattern.

Pathotype-specific resistance

Following basal resistance, the adapted pathogen encounters the second layer of host defence, race- or pathotype-specific resistance. This type of resistance is mediated by resistance (R) genes through the recognition of specific pathogen effectors. These effectors initially function as virulence/pathogenicity factors for the pathogen, but then become recognized by the host as signals of attempted infection (Bernoux et al. 2011). Most R genes encode proteins carrying a nucleotide-binding site (NBS) in the central region and a leucine-rich repeat (NBS-LRR) domain at the C-terminus. These genes form a large multigene family in the plant genome and can be separated into two subclasses, the toll-interleukin-1 (TIR) class and the coiled-coil (CC) class (Rafiqi et al. 2009). The NBS–LRR proteins are thought to recognize effectors or effector-specific signals from the pathogen and activate the plant immune system to provide hypersensitive responses (Bernoux et al. 2011).

Although many clubroot resistance (CR) loci have been identified through genetic analysis and quantitative trait loci (QTL) mapping (Piao et al. 2009), the resistance mechanisms associated with these genes have remained unknown until recently, when two R genes were cloned and characterized. The first gene is CRa (Ueno et al. 2012) from B. rapa, which confers specific resistance to P. brassicae isolate M85. Fine mapping of the CRa locus to the A. thaliana and B. rapa genomes revealed a candidate gene encoding a TIR-NBS-LRR protein. The cloned alleles of this gene in susceptible and resistant B. rapa lines have several structural differences, and CRa expression was observed only in the resistant line. Four mutant lines lacking clubroot resistance were obtained by the UV irradiation of pollen from a resistant line, and all of these mutant lines were found to carry independent mutations in this candidate gene. This genetic and molecular evidence strongly suggests that the identified gene is CRa.

The second gene is Crr1a (Hatakeyama et al. 2013), which is also from B. rapa. Crr1a is one of the two loci of Crr1 that was previously considered as a single locus. This gene (named Crr1a^G004) was cloned from the resistant line G004 and encodes a TIR-NB-LRR protein. By comparison, the susceptible allele Crr1a^A9709, cloned from the same locus in the susceptible line A9709, encodes a truncated NB-LRR protein. The Crr1a^G004 protein is expressed in the stele and cortex of the hypocotyl and roots, but not in root hairs, suggesting that it controls resistance to secondary infection or later pathogen development. Gain-of-function analysis proved that Crr1a^G004 was sufficient to bestow resistance to isolate Ano-01 in susceptible Arabidopsis and B. rapa genotypes.

The above two studies provide a basis for further molecular analysis of defence mechanisms against P. brassicae and will contribute to the breeding of resistant cultivars using marker-assisted selection. Furthermore, since resistance genes from different plant genomes often show similar structures because they are most likely to be derived from a common ancestral gene (Leister 2004), the information about CRa and Crr1a will facilitate the cloning and molecular characterization of other clubroot resistance genes.

Conclusions and the way forward

The molecular basis of clubroot pathogenesis is still largely unknown. Genomic studies of P. brassicae are still in their infancy, and the release of the whole genome sequence of the pathogen is highly anticipated. The fact that very few data can be referenced from other plant pathosystems, because of the unique taxonomic position of P. brassicae, suggests the need to conduct more basic molecular studies. The appropriate deployment of new techniques should be encouraged to maintain the pace of molecular studies. Although A. thaliana has proven to be an excellent model plant for the study of pathogenicity and host resistance in the clubroot pathosystem, more sequence data on the host genomes and more studies on the crop hosts would be extremely valuable, given the high demand for developing resistance in brassica cultivars of canola, European brassicas, Chinese cabbage and other Oriental crops. Based on the accumulated data from resistance gene/QTL mapping, studies on candidate genes will progress to the molecular level. Meanwhile, more resistance genes with diverse origins and mechanisms of function need to be identified. Increasing the diversity of resistance genes may allow for the partition of resistance according to at least the two distinct infection stages.

Acknowledgements

The authors gratefully acknowledge financial support from the Alberta Crop Industry Development Fund, the Alberta Canola Producers Commission, SaskCanola, the Manitoba Canola Growers Association, and from the Agriculture and Agri-Food Canada/Canola Council of Canada Clubroot Risk Mitigation Initiative.