Roles of Iron in Plant Defence and Fungal Virulence

Abstract

Iron is an essential component of various proteins and pigments for both plants and pathogenic fungi. However, redox cycling between the ferric and ferrous forms of iron can also catalyse the production of dangerous free radicals and iron homeostasis is therefore tightly regulated. our work has indicated that monocot plants challenged by pathogenic fungi redistribute cellular iron to the apoplast in a controlled manner to activate both intracellular and extracellular defences. In the apoplast, the accumulation of free, reactive ferric iron mediates defensive H₂O₂ production. Inside the cell, this efflux of iron creates a state of iron depletion, which directs the transcription of pathogenesis-related genes in concert with H₂O₂. In this addendum, we describe differences between the roles of iron in mediation of the oxidative burst in cereal and Arabidopsis responses to fungal pathogens. Also, we discuss the implications of current work concerning fungal iron uptake on host defence strategies.

Iron and the Oxidative Burst in Plants

The oxidative burst, a localized release of reactive oxygen species (ROS) by host plant cells following pathogen recognition, is well-documented.^1,2 In the model plant Arabidopsis thaliana, the respiratory burst oxidase homologues (RBOH) AtrbohD and AtrbohF are responsible for the majority of hydrogen peroxide (H₂O₂) production following inoculation with the avirulent bacterium Pseudomonas syringae pv tomato or the virulent oomycete Hyaloperonospora parasitica.³ Since the RBOHs are NADPH oxidases that produce superoxide (O₂·⁻), this H₂O₂ arises via dismutation rather than being a direct product of the RBOHs. Likewise, Nicotiana benthamiana NbrbohA and NbrbohB, which are similar to AtrbohF and AtrbohD, respectively, are required for H₂O₂ production in response to Phytophthora infestans.⁴ In addition to RBOHs, type III peroxidases have recently been shown to generate significant levels of H₂O₂ in Arabidopsis suspension cultures in response to pathogen-derived elicitors.⁵ Significantly, silencing of peroxidises, but not RBOHs, led to increased Arabidopsis susceptibility.^3,5 In monocots, the role of individual ROS generators is less clear. In wheat and barley, H₂O₂ is produced at defensive cell wall appositions (CWAs); fortified papillae of crossed linked phenolics, suberin, callose and proteins.⁶ O₂·⁻ however, is produced in epidermal cells only in association with successful fungal penetration.^7,8 Recently, Trujillo et al⁹ found that silencing the barley AtrbohF homologue HvrbohA led to increased penetration resistance against the powdery mildew fungus Blumeria graminis f. sp. hordei, suggesting that O₂·⁻ is required for cellular accessibility in that system. Unlike O₂·⁻, H₂O₂ is produced at wheat and barley CWAs in response to successful or defeated host fungi as well as nonhost fungi, and is therefore linked to basal resistance, which is active in all plants against all pathogens.

Basal resistance-linked H₂O₂ production can be seen as early as 3 h after inoculation, and in the wheat-powdery mildew pathosystem it is always associated with the accumulation of ferric iron.¹⁰ This finding shows interesting parallels between iron-mediated H₂O₂ production in plants and animals. Iron accumulation at CWAs is not specific to wheat; we found similar CWA-associated ferric iron in pathogen challenged barley, oat, corn, sorghum and millet, suggesting that iron accumulation is a universal phenomenon in cereals. This CWA-associated iron is important in the context of the oxidative burst because when the iron is blocked from accumulating, either with the iron chelator deferoxamine (DFO) or by blocking iron-laden vesicle like bodies from arriving at CWAs with the actin filament disruptor cytochalasin A, there is a concomitant loss of H₂O₂ production at the CWA. While basal resistance in both cereals and Arabidopsis involves the elaboration of a CWA and H₂O₂ production,¹¹ the source of the H₂O₂ differs. We have now shown that basal resistance-linked H₂O₂ production is mediated by CWA iron accumulation in cereals, but we found no iron accumulation at Arabidopsis CWAs (Fig. 1). The broad loss of resistance to both virulent and avirulent pathogens in peroxidase-silenced Arabidopsis⁵ suggests that peroxidase may fill the role that iron plays in cereals. Interestingly, the Arabidopsis AtrbohD-dependent oxidative burst is activated by ROS and goes on to limit the spread of pathogen-induced cell death,¹² suggesting that the knee-jerk production of H₂O₂ by peroxidases may prime a more targeted ROS production pathway. It would be interesting to investigate a possible interaction between iron-mediated H₂O₂ production and RBOH-dependent cellular accessibility in cereals.

Figure 1.

Differential iron accumulation at wheat and Arabidopsis cell wall appositions (CWA). A blue ring of ferric iron (Fe³⁺) can be seen surrounding the Blumeria graminis f. sp. tritici penetration attempt on the wheat epidermis (top), but is absent from the attack site on the Arabidopsis leaf (bottom). Fixed leaves were stained for iron using the Prussian blue technique. C, conidium; app, appressorium.

Iron and Fungal Pathogenesis

Having found important roles for iron in cereal defence mechanisms, we also began to investigate the role of iron uptake in the fungal pathogen Fusarium graminearum (Greenshields et al., In Press). In order to scavenge host iron for survival, fungal pathogens have evolved at least two iron acquisition systems. One system is hinged on the secretion and subsequent uptake of ferric iron-specific siderophores and the other system uses cell wall iron reductases to free bound or insoluble ferric iron by reducing it to ferrous iron for uptake.¹³ Recently, Oide et al.¹⁴ described the role of the nonribosomal peptide synthase NPS6 in extracellular siderophore production and showed that it was required for full virulence in the ascomycetes Cochliobolus heterostrophus, C. miyabeanus, F. graminearum and Alternaria brassicicola. Interestingly, siderophore production is not required for virulence in the basidiomycetes Ustilago maydis¹⁵ and Microbotryum violaceum,¹⁶ but loss of the ferroxidase/permease system of reductive iron uptake leads to a reduction in U. maydis virulence.¹⁷ On the surface, it seems convenient to attribute these different modes of infection-related iron uptake to the taxonomic distance between ascomycetes and basidiomycetes. However, neither Saccharomyces cerevisiae nor Candida albicans can produce or secrete siderophores¹³ and B. graminis f. sp. hordei spores show abundant ferric reductase activity.¹⁸ Also, B. graminis f. sp. hordei conidia express the multicopper oxidase gene FET3,¹⁹ which is required for full virulence in U. maydis.¹⁷ In light of our data showing the induction of wheat PR genes following DFO treatment, it is tempting to consider that pathogen produced siderophores may in fact work as pathogen-associated molecular patterns (PAMPS) in triggering host defences. All known plant pathogenic fungi that require siderophore production for virulence are necrotrophs,¹⁴ which are not as sensitive to recognition by the host as biotrophs.²⁰ Also, U. maydis, the only biotrophic pathogen characterised with respect to iron uptake, uses the reductive uptake system, thus avoiding secretion of potentially recognisable siderophores. Similarly, what little evidence that exists suggests that biotrophic B. graminis f. sp. hordei also uses the reductive iron uptake system.^18,19 It will be interesting to explore this relationship further and to understand how fungal siderophores are recognised and handled by host plants.

Addendum to: Liu G, Greenshields DL, Sammynaiken R, Hirji RN, Selvaraj G, Wei Y. Targeted Alterations in Iron Homeostasis Underlie Plant Defense Responses. J Cell Sci. 2007;120:596–605. doi: 10.1242/jcs.001362.