Several wall-associated kinases participate positively and negatively in basal defense against rice blast fungus

Plants have evolved the ability to detect potentially pathogenic microorganisms via pattern-recognition receptors (PRRs) localized on the surface of plant cells [1]. PRR proteins recognize Pathogen Associated Molecular Patterns (PAMPs) that are conserved motifs in the pathogen and Damage Associated Molecular Patterns (DAMPs) that derive from the damages caused by pathogen ingress [2]. Detection of pathogen through PRRs triggers PAMP-triggered immunity (PTI, also called basal defense) which is accompanied with rapid production of reactive oxygen species (ROS), activation of mitogen-activated protein kinases (MAPKs) and changes in expression of immune-related genes [2].

So far eight bacterial, four fungal PAMPs and 20 PRRs have been identified molecularly [3]. The best studied PAMP recognition systems in plants are represented by the bacterial flagellin recognized by the Arabidopsis thaliana FLS2 receptor and the fungal chitin recognized by the CEBiP receptor [1]. The FLS2 protein belongs to the Receptor-like Kinase (RLK) gene family. The typical structure of an RLK is an extracellular receptor domain that recognizes the PAMP molecule, a transmembrane domain and an intracellular kinase domain [4]. The CEBiP protein is composed of an extra-cellular LysM domain anchored to the membrane but does not contain any kinase domain [5]. FLS2 and CEBiP are found associated with RLK proteins like BAK1 in Arabidopsis and CERK1 in rice respectively [1]. FLS2 and CERK1 are positive regulators of basal defense since mutations in these genes lead to a decrease of resistance in Arabidopsis [6, 7] or to a decrease of basal defense in rice [8]. By contrast to PAMP, our knowledge on DAMP detection is much less advanced and only three pairs of PRRs and DAMP have been identified so far [3]. One of these is the PRR/DAMP pair between the Arabidopsis Wall-Associated Kinase 1 (AtWAK1) and oligogalacturonides (OGs) [9] derived from the pectin embedded in the cell-wall of most plants [10].

Wall-Associated Kinases are characterized by an extracellular domain composed of one or several repeats of the Epidermal Growth Factor (EGF) domain. The EGF domain is known in animals to bind a very large range of small peptides and to dimerize upon calcium binding [11]. EGF- containing proteins can form homo and heterodimers after ligand binding in animals [12]. Based on homology with the kinase domain of five WAKs from Arabidopsis [13], 21 genes coding WAK-like (WAKL) proteins were identified in Arabidopsis and 125 in rice, revealing an expansion of the WAK family in monocots [14, 15]. For simplicity and following previous nomenclature in rice [15], the WAK-like proteins are referred as WAKs. Among the rice WAKs, 67 have a bona fide EGF extracellular domain. Only a few WAKs from Arabidopsis or rice have been shown to possess kinase activity [16, 17]. Similarly, only a few WAKs have been localized to the plasma membrane in Arabidopsis [18] or rice (OsWAK1) [17], (OsDEES1/OsWAK91) [19]. More recently, maize ZmWAK was shown to be localized to the plasma membrane [20]. Moreover, WAKs seem to be found in large membrane protein complexes of unknown composition [21]. It is not known whether WAKs associate with other RLKs to ensure appropriate function like several other RLKs [22].

In plants, several ligands were shown to bind the extracellular domain of WAK proteins. For example the AtGRP3 protein binds to AtWAK1 [21] and pectin and OGs bind AtWAK1 and AtWAK2 [23–25]. It was shown that upon pectin treatment AtWAK2 activates the mitogen-activated kinases MPK3 and MPK6 and that a TAP-tagged (Tandem Affinity Purification) version of AtWAK2 constitutively activates ROS production and defense gene expression [26]. However, there is no indication that native WAKs can trigger ROS and there is only very limited information on defense gene expression during infection [20].

WAKs are involved in plant development [27]. For instance, AtWAK1 and AtWAK2 are required for cell wall expansion [28]. Accordingly, WAK mutants are often affected in their development. In rice, plants silenced for OsDEES1/OsWAK91 displayed fertility deficiency [19] that was attributed to a defect in embryo development. Plants silenced for the rice indica OsiWAK1 gene were stunted [29] and in Arabidopsis, silencing of AtWAK1 and AtWAK2 is lethal [28].

The role of WAKs in plant disease resistance initially came from indirect evidence with WAK mutants affected in the triggering of defense-related response [18]. Later, several studies provided direct evidence that WAK genes participate to resistance. First, it was shown that the RFO1/WAKL22 gene is responsible for quantitative resistance to Fusarium [30] and Verticilium [31]. More recently, two distinct wall-associated kinases from maize were shown to be responsible for a major QTL for resistance to the soil-borne fungus Sporisorium reilianum (ZmWAK) [20] and one against the foliar fungal pathogen Exserohilum turcicum (Htn) [32]. Secondly, several mutant analyses of WAK genes provided evidence for their involvement in disease resistance. The over-expression of AtWAK1 led to enhanced resistance to Botrytis [9] and over-expression of OsWAK1 enhanced resistance to Magnaporthe oryzae [17]. On the other hand, silencing of SlWAK1 in tomato lead to enhanced susceptibility to the bacterial pathogen Pseudomonas synringae pv tomato [33]. Other examples of the effect of WAKs on bacterial and fungal resistance are reported although the corresponding proteins miss an EGF domain (OsWAK25) [34] or a kinase domain (At5g50290) [35]. Thus several WAK mutants seem to act as positive regulators of disease resistance to fungi and bacteria without visible developmental phenotypes. However, there is thus far no indication that PTI is affected in these mutants.

Another indication that WAKs are related to disease response comes from the observation that WAK genes are often regulated by bacterial infection in Arabidopsis [33] and by blast infection in rice [36, 37]. Quite interestingly, there are two cases of pathogens that manipulate WAK gene expression by either expressing small RNA interfering with their RNA [35] or by an unknown mechanism [33]. Thus WAKs are important components of basal defense that pathogens try to inhibit. PAMPs can also directly regulate the expression of WAK genes [38]. Flagellin induces several WAK genes in Arabidopsis [39] and tomato [33]. Chitin induces OsWAK91 in rice in a CEBiP dependent manner in cell cultures [5] and the AtWAKL10 gene in Arabidopsis [40]. However, the global regulation of WAK genes in PTI is not well understood.

Here we report that several rice WAK genes are up-regulated while OsWAK112d is down-regulated by fungal infection in rice. Part of this transcriptional control is likely due to chitin detection by the chitin receptor CEBiP. We provide evidence that OsWAK14, OsWAK91 and OsWAK92 act as positive regulators of quantitative resistance, while OsWAK112 acts as a negative regulator. By studying OsWAK91 mutants, we demonstrate that this WAK significantly participates to ROS production and defense gene expression during infection.