Recognition at the leaf surface


(tel +48 618528503;


Recognition of interacting partners in plant–microbe relationships has, so far, mainly been discussed in the context of symbiotic interactions or pathogen detection by the plant immune system, but chemical recognition is also important for pathogens to detect their host and initiate the infection process. The study by Weis et al. in this issue of New Phytologist (2014, pp. 1310–1319) significantly contributes to our understanding of this interesting and likely complex phenomenon. In addition to broadening our knowledge of the molecular mechanisms of sensing plants by pathogens, the results of Weis et al. (2014) point at the complexity of plant metabolic pathways and the ambiguity of biochemical phenotypes of particular biosynthetic mutants.

‘Among the unanswered questions remains the role of aliphatic aldehydes in the prepenetration process.’

Weis et al. (2014) addressed the molecular basis of the earlier reported enhanced resistance of Arabidopsis cyp83A1 knockout plants to the adapted powdery mildew fungus Erysiphe cruciferarum (Weis et al., 2013). As indicated in the current study these mutants support, to a significantly lower degree than wild type plants, several stages of powdery mildew development including spore germination, appressoria differentiation and conidiophore formation. CYP83A1 is a cytochrome P450 monooxygenase that converts methionine-derived oximes into the respective thiohydroximates during the biosynthesis of aliphatic glucosinolates (AGs; Naur et al., 2003). Glucosinolates are sulfur-containing secondary metabolites present in the order Brassicales with multiple functions in plant adaptation to the environment, particularly in defense against insects and microbial pathogens (Hopkins et al., 2009; Bednarek, 2012). Products of glucosinolate hydrolysis are deterrents against generalist insect herbivores, while specialist insects have evolved different molecular mechanisms to avoid these toxic compounds. Remarkably, many insects specialized on Brassicaceae plants use glucosinolates, or their metabolic products, as an obligatory oviposition or feeding cue (Hopkins et al., 2009). Reduced spore germination and appressoria differentiation of E. cruciferarum on cyp83A1 plants suggested that this adapted pathogen could also use AGs or related compounds to identify its host. However, as indicated by Weis et al. (2014) the myb28 myb29 mutant, another AG-deficient Arabidopsis line (Hirai et al., 2007), showed a wild type-like phenotype upon inoculation with E. cruciferarum spores, disproving the hypothetical contribution of AGs to this interaction. Interestingly, AG-deficiency is not the only biochemical phenotype of cyp83A1 mutants that were also found to be strongly affected in the accumulation of sinapic acid and its derivatives (Fig. 1; Hemm et al., 2003). These metabolites represent phenylanine-derived phenylpropanoids that are not biosynthetically linked to AGs and it was shown that oximes, which are CYP83A1 substrates, can inhibit in vitro activity of caffeic acid O-methyltransferase (COMT), which catalyzes the final step of sinapic acid formation (Hemm et al., 2003). Metabolic analysis carried out by Weis et al. (2014) revealed elevated levels of 5-methylthiopentanaldoxime, an intermediate in the biosynthesis of 4-methylsulfinylbutylglucosinolate, and its isomer in the cyp83A1 mutant (Fig. 1). This clearly indicated that elimination of an oxime-metabolizing enzyme leads to in planta hyperaccumulation of its substrate(s), which in turn can impact other metabolic pathways, for example, sinapic acid biosynthesis. In addition, as oximes are characterized with relatively high chemical reactivity and biological activity, hyperaccumulation of such compounds could affect resistance of cyp83A1 plants towards E. cruciferarum.

Figure 1.

Chemical structures of compounds discussed in the text.

In addition to elevated oxime levels, Weis et al. (2014) revealed another, rather unexpected, phenotype of cyp83A1 plants: enhanced cuticle permeability. Consisting of cutin and waxes, the plant cuticle is supposed to function in plant immunity as the outermost physical barrier (Yeats & Rose, 2013). In addition, cutin fragments generated by cutinases produced by pathogenic microorganisms are considered as damage-associated molecular patterns (DAMPs), which are potent to elicit plant defense responses (Fig. 2). But defects in cuticle formation do not necessarily have to negatively correlate with resistance, for instance, Arabidopsis cutin-deficient mutants appeared to be more resistant to some necrotrophic fungal pathogens including Botrytis cinerea and Sclerotinia sclerotiorum (Bessire et al., 2007; Tang et al., 2007). This increase in resistance correlated with enhanced diffusion of a fungal elicitor(s) through the more permeable cuticle, which in turn induced enhanced secretion of a yet unidentified antimicrobial metabolite(s) (Fig. 2) (Bessire et al., 2007). By contrast, cutin deficient mutants did not show any alteration in their susceptibility towards powdery mildews suggesting that the proposed resistance mechanism does not affect this class of biotrophic pathogens (Bessire et al., 2007; Tang et al., 2007). To address the molecular basis of the observed increased cuticle permeability, Weis et al. (2014) performed analysis of cyp83A1 cuticular wax composition. This revealed a significant reduction of total wax content combined with changes in wax composition compared with wild type plants. Particularly striking was the absence of long chain aldehydes in the wax fraction isolated from cyp83A1 leaves. Notably, these particular metabolites have already been shown to initiate in vitro conidia germination, nucleus migration and appressorium differentiation of the barley powdery mildew Blumeria graminis f.sp. hordei (Hansjakob et al., 2012). The importance of these compounds for the prepenetration phase of powdery mildew development was further supported by significantly reduced B. graminis f.sp. hordei conidia germination and appressorial differentiation in the wax deficient maize glossy11 mutant (Hansjakob et al., 2011). Using chemical complementation of cyp83A1 plants with n-hexacosanal (Fig. 1), Weis et al. (2014) proved that lack of long chain aldehydes on the leaf surface was responsible for reduced germination of E. cruciferarum on cyp83A1 leaves. This finding highlighted the significance of chemical recognition as the trigger of prepenetration development. Of note, this phenomenon seems to be not only restricted to the powdery mildew pathogen. Analysis of the Medicago truncatula mutant inhibitor of rust germ tube differentiation1 revealed that epicuticular wax is critical for spore differentiation of two rust pathogens, Phakopsora pachyrhizi and Puccinia emaculata, and the anthracnose fungus Colletotrichum trifolii (Uppalapati et al., 2012).

Figure 2.

Cuticle as a double-edged sword in plant–microbe interactions. Beneficial contribution of the cuticle to plant immunity is indicated in green; advantages brought to pathogens indicated in red (Bessire et al., 2007; Hansjakob et al., 2012; Uppalapati et al., 2012; Yeats & Rose, 2013; Weis et al., 2014, this issue of New Phytologist, pp. 1310–1319). DAMP, damage associated molecular pattern, MAMP, microbe associated molecular pattern.

Overall, the results of Weis et al. (2014) indicated that the enhanced resistance of cyp83A1 plants to E. cruciferarum was based primarily on the deficiency of long chain aldehydes in the cuticular wax, which are apparently critical to initiate fungal development. In addition, the observed hyperaccumulation of putatively toxic oximes in leaves can further increase resistance of cyp83A1 plants. Among the unanswered questions remains the role of aliphatic aldehydes in the prepenetration process. As suggested by Weis et al. (2014) and other researchers (Hansjakob et al., 2012) these compounds likely act as chemical cues for the induction of spore germination and differentiation of infection structures. It would be intriguing to know if aliphatic aldehydes are simply necessary to inform the pest about appropriate conditions for the initiation of development or if they have any additional functions. In this context, it could be of significance that during their evolution, powdery mildews lost certain enzymes or even entire primary metabolic pathways and in the case of corresponding metabolites are likely completely dependent on their acquisition from the host (Spanu et al., 2010). It is not clear at the moment whether the pathways for long chain aliphatic aldehyde biosynthesis are intact in E. cruciferarum and other powdery mildews. It cannot be excluded that some of these compounds, although not produced endogenously, are necessary for these pests for some molecular mechanism underlying the early developmental stages. In such a case acquisition of aliphatic aldehydes from the plant cuticle could have additional significance for powdery mildews.

The other remaining question concerns the link between CYP83A1 function and deficiency of long chain aldehydes in cuticular wax. One of the options mentioned by Weis et al. (2014) could be a direct involvement of this enzyme in fatty acid metabolism. The CYP83 family belongs to the huge CYP71 clade of cytochrome P450 monooxygenases, which represents more than half of all CYPs in higher plants (Nelson & Werck-Reichhart, 2011). However, only a few members of the CYP71 clan have been proven to be linked with fatty acid metabolism; these include CYP77s, CYP703s and CYP92B1, which are not very closely related to CYP83A1 (Nelson & Werck-Reichhart, 2011). Oxime-metabolizing CYP83 enzymes were postulated to evolve in the order Brassicales from CYP71 enzymes converting oximes to α-hydroxynitriles in cyanogenic glucoside biosynthesis (Hansen et al., 2001). Experimental evidence suggested that substrate specificity of oxime-converting CYP71 and CYP83 monooxygenases is primarily determined by the presence of the oxime group (Naur et al., 2003). So far there is no evidence for oximes as intermediates in the formation of cuticular wax constituents (Yeats & Rose, 2013). For these reasons direct involvement of CYP83A1 in wax biosynthesis is not very likely. Another possible explanation for the reduced wax content in cyp83A1 plants could be a negative impact of the hyperaccumulating oximes on the activity of an enzyme(s) involved in wax biosynthesis, similar to the inhibition proposed for COMT (Hemm et al., 2003; Weis et al., 2014). In any case, cuticular wax deficiency in an AG-related mutant highlights the complexity of plant metabolic pathways.


Work in the P.B. laboratory is supported by an EMBO Installation grant and the National Science Centre SONATA BIS (UMO-2012/07/E/NZ2/04098) grant.