Numerous plant disease resistance (R) genes to viruses, bacteria, oomycetes, fungi and nematodes have been cloned over the last few years, with an overrepresentation of genes of the model species Arabidopsis thaliana (Gururani et al., 2012). The vast majority of these encode cytoplasmic proteins of the nucleotide binding-leucine rich repeat (NB-LRR) superfamily, and nearly all R genes operating against bacteria and oomycetes are cytoplasmic NB-LRR (Gururani et al., 2012). By contrast, R genes towards fungal pathogens are much more varied possibly because of the extreme diversity of fungal phytopathogens and of their modes of life within the plant tissues.
‘As in other models there is a need to reconcile terminology and there will be debate over whether the cloned gene is Rlm1 or LepR3 until Rlm1 is actually cloned.’
Surprisingly, although several disease resistance genes have been cloned in a few crop species (wheat, barley, potato) none has previously been in oilseed rape (Brassica napus), in spite of its relatively smaller genome and its close taxonomic proximity with A. thaliana. In this issue of New Phytologist Larkan et al. (pp. 595–605) provide the first report of the cloning of an R gene from B. napus, LepR3, operating against Leptosphaeria maculans, a fungal pathogen of major international importance. Departing from a resynthesized B. napus, Larkan et al. produced backcross generations of B. napus segregating for this R gene to map the resistance locus. After confirming collinearity of the genomic region with A. thaliana, and the recently sequenced Brassica rapa, the authors created targeted markers with which the position of the LepR3 locus was revised. The cloning of LepR3, orthologous to AtRLP32 (Wang et al., 2008), revealed that the predicted LepR3 protein is a receptor-like protein (RLP), similar to that found for the Cf R genes in the model system Cladosporium fulvum–tomato. Larkan et al. used transgenic L. maculans with the AvrLm1 avirulence gene (recognized by the Rlm1 resistance gene) and LepR3–transgenic B. napus and showed that LepR3 ‘interacts’ with the avirulence effector AvrLm1. With Rlm1 resistance overcome in France within 3–4 yr of cropping of Rlm1 in relation to the dispensability of the AvrLm1 effector in the genome of L. maculans (Rouxel et al., 2003; Gout et al., 2007), this explains why LepR3 was rendered ineffective in Australia only a few years after it was deployed in widespread commercial production. This also questions the existence of the cognate AvrLepR3 gene, at least in Australian populations of the fungus.
One issue contributing to the difficulty in cloning resistance genes in oilseed rape has been that many of the genetic maps (and currently on-going efforts to clone R genes) have been (or currently are) developed by private companies; in addition, different research groups use different nomenclature for both linkage groups and resistance genes (Hayward et al., 2012). As in other models there is a need to reconcile terminology and there will be debate over whether the cloned gene is Rlm1 or LepR3 until Rlm1 is actually cloned. However, the resistance gene cloned by Larkan et al. has remarkable features and opens the way to more generic and interrelated questions regarding RLPs as a possible generic sensing system for apoplastic fungi (and thus the need to have a definite understanding of the fungal mode of life in planta) and the (in)adequacy of A. thaliana as a model to decipher Brassica napus–Leptosphaeria maculans interactions.
Plant innate immunity to pathogens is usually divided into two lines of active defense: the first line is formed by pattern recognition receptors (PRRs), which are cell surface receptors that recognize highly conserved molecules termed pathogen-associated molecular patterns (PAMPs; Jones & Dangl, 2006). The response to PAMP recognition is termed PAMP-triggered immunity (PTI). Successful pathogens produce secreted effectors that suppress PTI responses, and during plant–pathogen co-evolution, plants have responded to these effectors through the development of cytoplasmic R proteins, mostly of the NB-LRR superfamily, which recognize the presence or activity of effectors that activate the effector-triggered immunity (ETI; Jones & Dangl, 2006). The generic assumption that cell surface receptors recognize PAMP is further substantiated by study of the Ve1 resistance gene of tomato conferring resistance to Verticillium dahliae race 1 that was originally considered an effector-targeted R gene and now tends to be regarded as a PRR (Thomma et al., 2011). As underlined by Thomma et al. (2011), this illustrates ‘the blurred PTI–ETI dichotomy’ with conserved chitin-binding or chitin scavenger effectors (such as Avr4 and Ecp6 of C. fulvum) that may be considered also as PAMPs. By contrast, the LepR3 gene is a RLP recognizing the species- and even isolate-specific effector AvrLm1, as do Cf genes of C. fulvum, and the candidate HcrVf2 gene which confers resistance to avirulent isolates of Venturia inaequalis in apple trees (Wang et al., 2010). There have been debates in the literature on how to classify the lifestyle of L. maculans (and other related Dothideomycetes). Fungi are classified as obligate biotrophs (ex. Blumeria graminis on barley) or pure necrotrophs (ex. Botrytis cinerea) but most fungal species behave as hemibiotrophs, that is, more or less complex mixes of biotrophic and necrotrophic stages. As noted on p. 1015 of Oliver (2012) ‘… pathogens cannot be divided into three clearly delineated classes (biotrophs, hemibiotrophs and necrotrophs) but rather into a complex matrix of categories each with subtly different properties.’ Leptosphaeria maculans has a very complex life cycle, encompassing saprophytism, biotrophy, necrotrophy and endophytic symptomless life and is described as an hemibiotroph (Oliver, 2012) or even as a necrotroph (Spanu, 2012), whereas related Dothideomycetes, with at least part of their infection strategies showing similarities with that of L. maculans, are classified as necrotrophs (Phaeosphaeria nodorum), hemibiotrophs (Mycosphaerella graminicola) or biotrophs (C. fulvum) (Oliver, 2012; Spanu, 2012). Regardless of their classification, L. maculans, C. fulvum, V. inaequalis and V. dalhiae share the particularity to grow extracellularly without penetrating plant cells and it makes sense that RLP are part of the surveillance machinery adapted to extracellular pathogens that evolved to recognize isolate-specific effectors in addition to more generic and ancient PAMPs. Following the pioneering work of Larkan et al., the cloning of other resistance genes to L. maculans is in progress (Hayward et al., 2012) and will allow us to critically evaluate this hypothesis.
An Arabidopsis thaliana–Leptosphaeria maculans pathosystem has been established by the C. Dixelius group in the mid-2000s (Staal et al., 2006). In A. thaliana, one accession out of 168 showed symptoms reminiscent of susceptibility observed on B. napus. However, this was accompanied by a very late development of the fungus compared with timing of symptom development, and compared with that observed in B. napus in which symptoms were expressed after extensive colonization of plant tissues, suggesting other factors (possibly toxins?) induced symptoms in the susceptible accession of A. thaliana. Staal et al. (2006) then identified a locus (AtRLM1Col) in Col-0 that comprised seven structurally related NB-LRR genes, of which two conferred increased resistance in the susceptible background while another locus in Ler-0 contained a paralogue. As underlined by Elliott et al. (2008), genes at this locus belong to a NB-LRR subgroup that does not seem to be under positive selection in A. thaliana, in contrast to classical R genes, which are constantly evolving in concert with their cognate Avr genes in the pathogen. Indeed, on the basis of characteristics of NB-LRR genes at the AtRLM1Col locus, and the paralogue gene, Jones & Dangl (2006) classified A. thaliana as a non-host for L. maculans. Consistent with its non-host status, and because the fungus cannot develop its typical endophytic growth in the tissues, A. thaliana resistance to L. maculans resembles that to a necrotrophic pathogen (Sasek et al., 2012). Actually, resistance responses towards L. maculans are mediated by jasmonic acid (JA) in A. thaliana, a feature characteristic of responses to necrotrophs, whereas the AvrLm1–Rlm1 interaction induces salicylic acid (SA) and ethylene (ET) signaling in B. napus, a feature characteristic of responses to biotrophs (Sasek et al., 2012). Interestingly, although generally considered as an unusual feature, the simultaneous activation of ET and SA pathways is also observed in the Avr2–Cf-2 and Avr9–Cf-9 Cladosporium fulvum–tomato interactions (Hammond-Kosack & Jones, 1996). In addition, none of the 57 RLP genes predicted in the A. thaliana genome was found to be involved in pathogen response when analyzing 14 plant pathogens, including nine fungi (but not L. maculans), with diverse lifestyles (Wang et al., 2008). This, however, would not be surprising if the orthologue of AtRLP32 in B. napus, LepR3, acts in a race-specific manner rather than as a PAMP sensor. These data illustrate the limits of A. thaliana as a model to decipher the complexity of interactions and resistance mechanisms set up during million years of plant–pathogen co-evolution. Brassica now enters the age of genomics with all main Brassica crops having been (or in the process of being) sequenced (http://www.brassica.info/). Following obtainment of a reference genome sequence for L. maculans, and major advances in the understanding of the fungal genome (Rouxel et al., 2011), its plasticity, and its effector complement, the cloning of the first resistance gene to L. maculans in oilseed rape now paves the way for detailed studies on the genomics of resistance and for dissection of the evolutionary plant–pathogen adaptations in this especially complex system.