La fusariose des épis des céréales est due à un champignon pathogène (Fusarium spp.) qui entraine non seulement une perte directe de rendement en interrompant le remplissage des grains, mais également une perte indirecte par la production de mycotoxines, responsables de perturbations de procédés industriels et d'intoxications alimentaires. Parmi les différents moyens de lutte, la création de variétés résistantes semble la stratégie la plus efficace et la plus durable. Plusieurs variétés exotiques ont ainsi été identifiées comme résistantes mais présentent la particularité d'être difficilement cultivables sous nos latitudes. L'obtention de variétés résistantes adaptées à nos climats passe donc par une meilleure connaissance des mécanismes de résistance disponibles. La détection de QTL (Quantitative Trait Loci) chez le blé tendre a déjà permis de localiser certains des facteurs génétiques impliqués dans des mécanismes de résistance. Cependant l'utilisation de ces résultats en amélioration des plantes se heurte à deux contraintes majeures : la position des QTL reste le plus souvent imprécise et ces QTL, dérivant de populations biparentales, ne représentent qu'une fraction de la diversité génétique potentiellement intéressante pour ce caractère. L'objectif de la thèse était d'étudier la résistance du blé à la fusariose en réalisant une approche de génétique d'association. L'utilisation de génotypes faiblement apparentés et d'une densité élevée de marqueurs a permis de détecter des associations entre le polymorphisme génétique et des variations phénotypiques. Après avoir réalisé une synthèse sur le pathogène et sur l'approche par génétique d'association, le premier travail a été de réaliser une carte génétique fiable afin de pouvoir localiser les associations mises en évidence. Un panel de 195 variétés de blé tendre élites a été évalué pour leur résistance globale à la fusariose, pour leur résistance à la progression des symptômes et pour leur résistance à l'accumulation des mycotoxines. Des notations morphologiques (précocité, taille, aristation, nombre d'épillets par épi et extrusion des anthères) ont également été réalisées afin d'évaluer leur influence dans l'infestation. Ces variétés ont été génotypées avec 3016 marqueurs SNP, 200 marqueurs SSR et 1400 marqueurs DArT. Préalablement à la réalisation des tests d'association, l'étude de la structure a mis en évidence trois origines géographiques (française, anglaise et allemande) des variétés du panel. De même, l'étude du déséquilibre de liaison a montré que celui-ci était conservé sur une distance comprise entre 2 et 6 cM. L'étude d'association réalisée sur les notations d'infestation a permis d'identifier plusieurs zones associées, dont plusieurs sont colocalisées avec des Meta-QTL et/ou avec des zones identifiées avec les notations morphologiques. La densification en marqueurs de 6 zones associées avec les différentes notations d'infestation, localisées sur les chromosomes 1A, 1D, 2A, 2B, 5A et 7A, ont permis de confirmer ces régions, de restreindre fortement l'intervalle d'intérêt et de mettre en évidence de potentiels gènes candidats. Cette analyse a ainsi permis de mieux comprendre les mécanismes génétiques et morphologiques impliqués dans la résistance à la fusariose des épis. La décomposition en trois partie de cette résistance (résistance globale, résistance à la progression des symptômes, et résistance à l'accumulation des mycotoxines) montre l'existence de mécanismes génétiques indépendants et donc complémentaire à prendre en compte dans une stratégie de création de variétés résistantes à la fusariose des épis. / Fusarium head blight (FHB), caused by the fungal pathogen Fusarium spp., is a major cereals disease that not only cause direct yield losses, by interrupting grain filling, but also indirect quality losses by the production of mycotoxins, responsible of industrial processes disturbance and food poisoning. Most wheat breeding programs in FHB-affected areas of the world have been screening germplasm to identify sources of improved tolerance. Unfortunately, these sources of genetic resistance are often of exotic origin and not adapted to West European growing conditions. The selection of adapted varieties with improved tolerance therefore needs a better characterization of resistance mechanisms. Several QTL for FHB resistance in wheat have been identified in European germplasm, but the use of these information in marker-assisted selection is constrained by the precision of the QTL and the low diversity tested by using biparental population.The aim of the present study is to use association mapping, also called linkage disequilibrium mapping, to identify loci involved in the resistance to FHB. This method refers to the analysis of statistical associations between genotypes determined in a collection of individuals, and the phenotypes of the same individuals. A dense genetic map was compiled to localize precisely the association results. Resistance to F. graminearum was studied in a panel of 195 elite wheat varieties by the evaluation of three components: resistance to global infection, resistance to symptom progression and resistance to accumulation of mycotoxins. Morphological factors (plant height, heading date, awnedness, spikelets per ears, anther extrusion), known to influence resistance to FHB, were also recorded. All the varieties have been genotyped with around 3300 SNP markers, 200 SSR markers and 1400 DArT markers. We first investigated the structure of the panel, which could generate bias in the estimate of allele effects, if not included explicitly in the association models. We showed that the structure is based on geographical origin (French, German and UK). Study of the linkage disequilibrium (LD) showed an extent of LD between 2 and 6 cM. Results of association studies permitted to identify several loci for each of the evaluated components of resistance. Some of these loci colocalized with the results of the MetaQTL analysis and/or with loci associated with morphological traits. We selected more specifically 6 loci, located on chromosome 1A, 1D, 2A, 2B, 5A and 7A. Marker saturation of the regions, allowed to confirm the genome wide association results and to increase the accuracy of the loci of interest. This analysis allowed to better understand the many factors that influenced FHB resistance, whether genetic or morphological. Results show that the genetic mechanisms are independent between the three components and therefore, information obtained for each component are to be used complementary to create varieties with increased resistance to FHB.

La fusariose des épis est l’une de maladies les plus dommageables des céréales tempérées et est principalement causée par le champignon toxinogène Fusarium graminearum (Fg). Ces dix dernières années, de nombreuses études ont rapporté l’induction transcriptionnelle de gènes de la plante codant pour des cytochromes P450 (P450) en réponse l’infection par Fg. Les P450s constituent une famille enzymatique impliquée dans de nombreuses voies métaboliques, certaines avec des intérêts potentiels dans la résistance face aux maladies. Nous avons utilisé la petite graminée modèle Brachypodium distachyon (Bd) pour caractériser fonctionnellement le premier gène codant pour un P450 induit chez la plante au cours de la fusariose des épis par l’utilisation de lignées altérées dans la séquence ou l’expression du gène Bradi1g75310 codant le P450 BdCYP711A29. Nous avons montré qu’en plus d’être un facteur de sensibilité à la maladie, le gène Bradi1g75310 est impliqué dans une voie de biosynthèse hormonale chez Bd, celle des strigolactones (SLs). En effet, en plus de complémenter génétiquement les phénotypes aériens de la lignée mutante max1-1 d’Arabidopsis thaliana altérée dans le gène homologue MAX1 (AtCYP711A1), une lignée de Bd surexprimant Bradi1g75310 (lignée OE) exsude davantage d’orobanchol, une SL spécifique, que la lignée sauvage ou mutante. Une analyse préliminaire de l’impact direct de l’orobanchol sur la croissance de Fg semble indiquer une activation des étapes précoces du développement du champignon (germination) qui pourrait être à l’origine de l’induction plus rapide de gènes de défenses observée chez une lignée OE de Bradi1g75310. Nous avons également montré que les 4 paralogues de Bradi1g75310 chez Bd, qui codent également pour des CYP711A, sont tous capable de complémenter la lignée max1-1 et avons généré du matériel végétal fondamental pour la poursuite de l’étude de la diversification des SLs chez la plante monocotylédone modèle Bd. Au global, ce projet constitue une première étape dans la caractérisation de l’implication des P450 dans la réponse de la plante face à l’infection par Fg en plus de donner de nouveaux indices concernant le rôle des SLs dans les interactions plante-pathogène. Les résultats obtenus au cours de ce travail de thèse pourront permettre l’amélioration de caractères tant développementaux que de résistance à la fusariose chez les céréales cultivées. / Fusarium Head Blight (FHB) is one of the most important diseases of temperate cereals and is mostly caused by the toxin producing-fungus Fusarium graminearum (Fg). This last decade, several studies reported the transcriptional activation of cereal cytochrome P450-encoding genes (P450s) in response to Fg infection. P450s constitute an enzymatic family participating in very diverse metabolic pathways with potential interest for disease resistance. We used the model temperate cereal Brachypodium distachyon (Bd) to functionally characterize the first FHB-induced P450- encoding gene using Bd lines altered in the locus or gene expression of the Bradi1g75310 gene encoding the BdCYP711A29 P450. We showed that in addition to be a plant susceptibility factor towards the disease, the Bradi1g75310 gene is involved in the hormonal biosynthetic pathway of strigolactones (SLs) in Bd. Indeed, in addition to genetically complement the shoot phenotypes of the Arabidopsis thaliana mutant line for the homologous gene MAX1 (AtCYP711A1, max1-1 line), a Bd linewhich overexpresses the Bradi1g75310 gene (OE) exudes more orobanchol, a specific SL, compared to wild-type or mutant lines. Preliminary analysis of the direct impact of orobanchol on Fg growth suggests an activation of early fungal development (germination) likely to induce faster induction of defense-related genes during FHB, observed in Bradi1g75310 OE line. We showed that the four paralogs of Bradi1g75310 encoding BdCYP711A P450s are all able to genetically complement max1-1 line and provide important plant material for studying SLs diversification in the model monocot B. distachyon. Overall, this project constitutes a first step in the characterization of P450s involvement in plant response towards Fg infection in addition to give new evidences about the role of SLs in plant-pathogen interactions. Results obtained during this Ph.D. project will allow the improvement of both developmental and FHB-related traits in cereal crops.

Some specialist microbes can deploy a range of mechanisms to cause disease on one or more host plant species. To identify entirely new classes of pathogenicity and virulence factors, a bioinformatics-reverse genetics approach has been applied to a plant pathogen where near complete genomic sequence information was available. A genomic region was identified on chromosome 1 of the important cereal pathogen Fusarium graminearum that contains a significant grouping of homologues of known virulence genes. Targeted deletion of these genes revealed a role for the neutral trehalase (NTH1) and protein kinase A regulatory subunit (PKAR) genes in the rate of disease symptom spread by F. graminearum, in addition to the previously reported SNF1 Ser/Thr protein kinase and STE7 MAP kinase kinase genes. Subsequent investigation of further genes at this locality revealed the presence of a gene, here named Fusarium graminearum Contributor to Virulence 1 (FCV1), which represent a novel class of gene required for a full rate of symptom spread. Targeted deletion of FCV1 led to a reduced rate of disease development by F. graminearum on wheat ears and Arabidopsis floral tissue, but did not affect trichothecene mycotoxin production. The fcv1 deletion mutant also exhibits altered hyphal growth, reduced asexual sporulation and altered sensitivity to oxidative and osmotic stress. In the complemented strain, wild-type traits were completely or partially restored. This micro-region of < 40 kb containing these five important genes represents a novel type of gene cluster containing genes required for a full rate of disease development. This micro-region is located in a genomic region of low recombination, is highly conserved in three other Fusarium species, but is less conserved in other plant pathogenic species. The micro-region is not defined by a distinct GC content or coordinated gene expression patterns, nor is it flanked by highly repetitive sequences. This micro-region is distinct from the previously identified fungal and bacterial virulence gene clusters and the clustered biosynthesis-associated genes required to synthesis metabolites which contribute to pathogenicity. This method for novel disease development-contributing gene identification is applicable to any sequenced pathogen species.

Fusarium graminearum is one of the most problematic phytopathogens in US agriculture. This fungus causes head blight, foot rot, and damping off on wheat and barley. The infection lowers the grain yield and causes contamination of the grain product with mycotoxins. Effective control measures are lacking, and new fungicides that kill F. graminearum but remain safe and economical to use are needed. Newly synthesized aminoglycosides (JL22, JL38, JL39, JL40, NEOF004, NEOF005), classic aminoglycosides (amikacin, gentamicin, kanamycin A, kanamycin B, neomycin, and ribostamycin), and a lipopeptide, syringomycin E (SRE), were studied to determine their antifungal potential to control F. graminearum. Aminoglycosides are protein synthesis inhibitors that mainly target bacteria, but a few were recently observed to kill fungi. They consist of an aminocyclitol ring bound with two or more amino sugars. Novel aminoglycosides were recently synthesized using novel glycodiversification synthetic schemes involving the replacement of the original amino sugars with unusual amino sugars. SRE is an antifungal lipodepsinonapeptide produced by Pseudomonas syringae pv. syringae. This bacterium is an opportunistic pathogen in a wide range of plant species and produces several fungicidal lipopeptides. SRE forms pores on fungal plasma membrane and causes ion fluxes. An enhancement of its antifungal activity is reported in the presence of rhamnolipid surfactants. The antifungal activities of various aminoglycosides, SRE, and a SRE-rhamnolipids mixture were determined against F. graminearum by measuring in vitro minimum inhibition concentrations (MICs) and in planta lesion area and chlorosis development using a leaf infection assay protocol. It was determined that using Tween® 20 at 0.2 % (v/v) concentration in the leaf infection assay promotes lesion development by F. graminearum with minimum phytotoxicity. In vitro, SRE, SYRA, and synthetic aminoglycoside JL38 showed the best antifungal activities. With the in planta assay, all three antifungal agents prevented infection by F. graminearum. However, inconsistent phytotoxicities were observed with SRE and SYRA that were influenced by the Tween® 20 surfactant included in the leaf infection assay. How Tween® 20 induces these phytotoxic inconsistencies is not known.

The fungus Fusarium graminearum (Schwabe) causes Gibberella ear rot in maize, resulting in accumulation of harmful mycotoxins in the grain. Disease severity and pericarp/aleurone dehydrodiferulic acid content are negatively correlated. Furthermore, quantitative trait locus mapping (QTL) identified colocalization between QTL for both traits. A candidate gene approach was employed to identify the genes responsible for the observed colocalization. Candidate genes selected on the basis of their putative involvement in various aspects of cell wall DFA accumulation were mapped in silico using the maize genome sequence. Polymorphisms were discovered in putative genes and converted to molecular markers. The in silico mapping effort was successful in predicting map locations of the analyzed sequences, and the segregation of certain marker alleles could explain variation for Gibberella ear rot severity and pericarp-aleurone DFA content.

Three cultivators, Rabewerk, Kongskilde and Hiniker, and three intercrops, soybean, lupin and red clover + rye grass were investigated for their impact on the incidence of ear rot of corn, a common disease of maize in eastern Canada. Wheat was seeded in the corn rows to serve as an additional indicator of cultivation and intercropping effects on the pathogen. An artificial inoculum of F. graminearum that produced perithecia and ascospores was used to mimic natural inoculum. / In 1993 and 1994, the infection in the corn was not very severe and there were no differences between the treatments and the controls. / In 1993 and 1994, at both sites, wheat seeds from cultivation trials showed a tendency for greater disease incidence in the non-cultivated herbicide treatment when compared to any of the other cultivator treatments. Cultivators till the soil and bury corn residues, this action led to the destruction of some of the inoculum and a subsequent reduction of the disease incidence in the cultivated plots. / In the intercrop trial of 1993 and 1994, wheat infection was moderate to severe, except at L'Assomption in 1993, but no significant differences were observed among the treatments. It is believed that interplot interference, due to ascospores moving from one plot to the next, masked differences. / The results indicated that weed cultivation would have a negligible or no effect on the development of fusarium ear rot of corn in Quebec. (Abstract shortened by UMI.)

Fusarium graminearum causes Fusarium Head Blight, (one of) the most destructive cereal diseases in Canada. Yield loss, quality degradation and mycotoxin production make Fusarium a multifaceted threat. Regulated production of reactive oxygen species by Nox enzymes is indispensable for fungal pathogenesis. F. graminearum Nox mutant ∆noxAB produced equivalent mycotoxin but caused reduced virulence than wild-type. We hypothesized that Nox mediated redox signaling may participate in F. graminearum pathogenicity. Two-DE and gel-free biotin affinity chromatography, followed by LC-MS/MS analysis were employed for a comparative redox-proteomics analysis between wild-type and ∆noxAB to identify proteins oxidized by Nox activity. Total 35 proteins, 10 by 2-DE and 29 by gel-free system, were identified. 34% proteins participated in fungal metabolism, 20% in electron transfer reactions and 9% were anti-oxidant proteins. The findings suggested that Nox mediated thiol-disulfide exchange in proteins provide a switch for redox-dependent regulation of metabolic and developmental processes during induction of FHB.

Fusarium head blight (FHB), a fungal disease caused principally by Fusarium graminearum, and barley yellow dwarf (BYD) caused by BYD luteoviruses are two serious fungal and viral diseases of wheat resulting in high economic losses annually. Wuhan, a Chinese wheat cultivar resistant to FHB, and Maringa, a Brazilian cultivar tolerant to BYDV were inter-crossed and crossed with Roblin, a Canada western red spring wheat susceptible to both FHB and BYDV, to determine the genetic basis of resistance/tolerance and to combine the two traits. Four hundred ninety nine F1-derived doubled haploid (DH) lines were generated from reciprocal crosses using corn pollen-mediated DH technology. The DH lines and the parents were evaluated for disease symptoms, reduction in height and spike mass for BYD and for disease incidence, disease severity and Fusarium-damaged kernels for FHB in field and controlled environments. A subset (20/150) of the best performing DH lines from Wuhan/Maringa populations for both BYD and FHB were further evaluated. Plants were point inoculated with F. graminearum in greenhouse experiments, and macroconidial spray inoculations and spread of corn inoculum were used in field environments to evaluate FHB. BYDV inoculations were performed by placing ten to fifteen viruliferous aphids (Rhopalosiphum padi infected with BYDV-PAV isolate 9301PAV), at the one to two leaf stage for both greenhouse and field trials. The studies showed that both FHB and BYDV are quantitatively inherited. Transgressive segregants were observed and the broad sense heritability was high (0.90 to 0.97) for all traits evaluated. Results from independent testing of diseases on Wuhan/ Maringa populations showed fourteen DH lines were as, or more resistant than Wuhan for FHB and Maringa for BYDV tolerance and have combined both BYDV tolerance and FHB resistance. Identifying such lines facilitates the pyramiding of independent genes to obtain adequate levels of enduring resistance. A further experiment was conducted on the 14 lines by inoculating them with BYDV and F. graminearum successively on the same plant. Six out of 14 selected DH lines demonstrated high resistance to FHB and tolerance to BYDV. These six lines can be used in FHB/BYDV resistance/tolerance breeding programs.

Fusarium graminearum causes Fusarium Head Blight, (one of) the most destructive cereal diseases in Canada. Yield loss, quality degradation and mycotoxin production make Fusarium a multifaceted threat. Regulated production of reactive oxygen species by Nox enzymes is indispensable for fungal pathogenesis. F. graminearum Nox mutant ∆noxAB produced equivalent mycotoxin but caused reduced virulence than wild-type. We hypothesized that Nox mediated redox signaling may participate in F. graminearum pathogenicity. Two-DE and gel-free biotin affinity chromatography, followed by LC-MS/MS analysis were employed for a comparative redox-proteomics analysis between wild-type and ∆noxAB to identify proteins oxidized by Nox activity. Total 35 proteins, 10 by 2-DE and 29 by gel-free system, were identified. 34% proteins participated in fungal metabolism, 20% in electron transfer reactions and 9% were anti-oxidant proteins. The findings suggested that Nox mediated thiol-disulfide exchange in proteins provide a switch for redox-dependent regulation of metabolic and developmental processes during induction of FHB.

Fusarium head blight (FHB), a fungal disease caused principally by Fusarium graminearum, and barley yellow dwarf (BYD) caused by BYD luteoviruses are two serious fungal and viral diseases of wheat resulting in high economic losses annually. Wuhan, a Chinese wheat cultivar resistant to FHB, and Maringa, a Brazilian cultivar tolerant to BYDV were inter-crossed and crossed with Roblin, a Canada western red spring wheat susceptible to both FHB and BYDV, to determine the genetic basis of resistance/tolerance and to combine the two traits. Four hundred ninety nine F1-derived doubled haploid (DH) lines were generated from reciprocal crosses using corn pollen-mediated DH technology. The DH lines and the parents were evaluated for disease symptoms, reduction in height and spike mass for BYD and for disease incidence, disease severity and Fusarium-damaged kernels for FHB in field and controlled environments. A subset (20/150) of the best performing DH lines from Wuhan/Maringa populations for both BYD and FHB were further evaluated. Plants were point inoculated with F. graminearum in greenhouse experiments, and macroconidial spray inoculations and spread of corn inoculum were used in field environments to evaluate FHB. BYDV inoculations were performed by placing ten to fifteen viruliferous aphids (Rhopalosiphum padi infected with BYDV-PAV isolate 9301PAV), at the one to two leaf stage for both greenhouse and field trials. The studies showed that both FHB and BYDV are quantitatively inherited. Transgressive segregants were observed and the broad sense heritability was high (0.90 to 0.97) for all traits evaluated. Results from independent testing of diseases on Wuhan/ Maringa populations showed fourteen DH lines were as, or more resistant than Wuhan for FHB and Maringa for BYDV tolerance and have combined both BYDV tolerance and FHB resistance. Identifying such lines facilitates the pyramiding of independent genes to obtain adequate levels of enduring resistance. A further experiment was conducted on the 14 lines by inoculating them with BYDV and F. graminearum successively on the same plant. Six out of 14 selected DH lines demonstrated high resistance to FHB and tolerance to BYDV. These six lines can be used in FHB/BYDV resistance/tolerance breeding programs.