In an effort to broaden our understanding of external triggers influencing the DON production machinery of F. graminearum, the effect of strobilurin and triazole fungicides on DON production was investigated. Our results demonstrate that prothioconazole, a triazole fungicide, has good control capacities culminating in reduced vegetative radial outgrowth, a reduced conidial germination and a reduction of F. graminearum biomass. Triazoles are known inhibitors of the ergosterol biosynthesis in fungi and have been described for their good control capacities against Fusarium spp .
On the contrary, the strobilurin fungicide azoxystrobin was not able to induce a reduction in radial outgrowth, spore germination and fungal biomass. Strobilurin fungicides inhibit mitochondrial electron transport by binding the Qo site of cytochrome bc1 complex. Although the effectiveness of strobilurins against Fusarium spp. is doubtable, they have been reported to be effective against F. culmorum  Apparently, F. graminearum is very resistant to this type of fungicides. Resistance to strobilurin fungicides has been reported in many species to be associated with a single amino acid replacement at position 143 of the cytochrome b gene [26–28]. Although this mechanism was recently described in Microdochium nivale it has not yet been described in F. graminearum. We assume that the observed resistance is therefore possibly a consequence of the activation of a respiratory chain using an alternative oxidase (AOX) bypassing complexes III and IV in the cytochrome mediated pathway. Activity of this AOX mediates electron transfer directly from ubiquinol to oxygen. Kaneko and Ishii (2009) demonstrated that F. graminearum acts very rapidly upon strobilurin application by the activation of AOX whereas M. nivale, a fungal species susceptible to strobilurins, reacted slowly with a retarded moderate activation of this enzyme .
Since the generation of reactive oxygen species such as H2O2 is a hallmark of an oxidative stress response, extracellular H2O2 was measured upon fungicide application in an in vitro assay. Unexpectedly, application of strobilurin fungicides did not result in an increased extracellular H2O2 formation, which is at first sight, contradictory to previous findings by Kaneko and Ishii (2009) who found an increased production of H2O2 upon strobilurin application. However it is important to notice that in the present work the H2O2 released in the medium was measured whereas Kaneko and Ishii (2009) focused on intracellular H2O2. Remarkably, the application of sub lethal doses of prothioconazole or the combination of prothioconazole amended with fluoxastrobin resulted in a boosted H2O2 production as fast as 4 h after application. This prompt production disappeared at later time points. In addition, a clear induction of DON production was observed 48 h after application of sub lethal prothioconazole + fluoxastrobin concentrations. This induction of DON was confirmed in an in vivo experiment in which flowering wheat plants were infected with F. graminearum and subjected to a sub lethal dose of prothioconazole + fluoxastrobin. Previous work on F. culmorum demonstrated no or a negative effect of several strobilurins and triazoles on DON production  so the observed phenomenon of an increased DON production by F. graminearum induced by sub lethal concentrations of triazole fungicides might be a strain- or species-specific phenomenon.
It is tempting to speculate whether this accumulation of DON is the consequence of the preceding accumulation of H2O2 as such being the first link in a signalling cascade activated upon sub lethal triazole treatment. Although this key role of H2O2 is not unambiguously demonstrated in the present study, the amount of evidence is compelling: H2O2 precedes accumulation of DON, combined application of catalase (eliminating H2O2 from the medium) inhibited DON accumulation. In addition, the application led to a reduced activity of the triazole fungicide. Application of H2O2 to F. graminearum cultures led to a reduced germination and prompt induction of DON biosynthesis 4 h after H2O2 application. This additional experiment proves that H2O2 accumulation is necessary and sufficient to initiate DON production. The activation of the DON biosynthesis machinery by H2O2 is in concordance with previous observations by the group of Barreau [17, 19, 20] who demonstrated that exogenously applied H2O2 by repeated single or pulse-feeding resulted in accumulation of DON. However, these authors only monitored increases in DON at late time points such as 10 to 30 days after H2O2 application whereas we observe a clear prompt activation of DON production within hours. From a physiological point of view the effect of H2O2 during the initial germination events is logic and in line with the physiology of an in field F. graminearum infection: H2O2 is one of the key regulators in the plant defense system upon pathogen attack . Therefore, this molecule is encountered frequently and at early time points by the pathogen in the interaction with its host. Previous work by the group of John Manners demonstrated beautifully that DON itself can induce hypersensitive cell death and H2O2 during infection  and as such underpinning the interaction between both molecules.
Astonishingly, very low concentrations of H2O2 promoted conidia germination rate where a reduction was expected. We hypothesize that during germination events, very small amounts of H2O2 are beneficial and necessary in the primordial germination- and hyphal extension events. It is known that H2O2 is necessary in de novo synthesis of cell wall and membrane components during germination and hyphal extension. Indirect evidence for this was provided by Seong et al (2008) who observed high activities of the peroxisomes at the onset of spore germination  The need for basal H2O2 is subscribed by the observation that catalase treatment results in a reduced spore germination at very early time points in germination. In several independent studies, it was demonstrated that reactive oxygen species such as H2O2 are key players and crucial in the regulation of cell differentiation in microbial eukaryotes [32, 33]. In accordance with this, it was demonstrated that NADPH oxidases which generate reactive oxygen are decisive in fungal cell differentiation and growth in a model system using Neurospora crassa .
Taken together, these results not only reinforce the hypothesis that H2O2 can induce DON biosynthesis but also suggest that DON accumulation induced by sub lethal triazole application is mediated through an increased production or release of H2O2 into the medium rendering a physiological interface of H2O2 influencing DON production. It is tempting to speculate on the mechanistics behind these observations. We hypothesize that due to the inhibition of ergosterol biosynthesis by the application of triazole fungicides, an increased cell permeability results in the increased release of H2O2 in the medium which in turns activates the trichothecene biosynthesis machinery. Indeed, although H2O2 is a very reactive molecule which can diffuse freely across bio membranes, it has been shown in a Sacharomyces model system that organisms prevent H2O2 diffusion [35, 36]. This hypothesis is subscribed by accumulating indirect evidence in many other fungi. As such in Candida ergosterol depletion increases vulnerability to phagocytic oxidative damage . In Sacharomyces it was demonstrated using ergosterol knock out mutants that ergosterol depletion results in a changed biophysical property of the plasma membrane leading to an increased permeability towards H2O2.
Although beyond the scope of the present paper it is important to notice that triazole fungicides on their own can generate H2O2
in planta as an intermediate metabolite in plants through activation of antioxidant systems  generating as such a greening effect which results in a retardation of the senescence . The effect of this physiological induced H2O2
in planta on DON production by an invading F. graminearum is till now not studied and certainly needs more attention in the future.