AVERTISSEMENT

Ce document est le fruit d'un long travail approuvé par le jury de soutenance et mis à disposition de l'ensemble de la communauté universitaire élargie.

Il est soumis à la propriété intellectuelle de l'auteur. Ceci implique une obligation de citation et de référencement lors de l’utilisation de ce document.

D'autre part, toute contrefaçon, plagiat, reproduction illicite encourt une poursuite pénale.

Contact : [email protected]

LIENS

Code de la Propriété Intellectuelle. articles L 122. 4 Code de la Propriété Intellectuelle. articles L 335.2- L 335.10 http://www.cfcopies.com/V2/leg/leg_droi.php http://www.culture.gouv.fr/culture/infos-pratiques/droits/protection.htm Faculté des Sciences U.F.R. Sciences et Techniqlles Biologiq'lIes Ecole Doctorale Ressources Procédés Produits Environnement Département de Formation Doctorale Sciences Agronomiques et Forestières, Biologie et Ecologie, Biotechnologie

SCD UHP NANCY 1 Bibliothèque des Sciences Rue du Jardin BOt?illlC1Ue - cs 54601 Vr~GERS LES NANCY CED~~ Thèse

Présentée pour l'obtention du titre de

Docteur de l'Université Henri Poincaré, Nancy 1

en Biologie Végétale et Forestière

par

Benoit BARRÈS

Etude de la structuration génétique de populations de Melampsora larici-populina, agent de la rouille du peuplier, il différentes échelles spatiales

Soutenue publiquement le 15 novembre 2006 devant la commission d'examen

Mme Claire NEEMA Professeure, INA, Paris-Grignon Rapporteure Mme Tatiana GIRAUD Chargée de Recherche, CNRS, Orsay Rapporteure M. Michel CHALOT Professeur, Université Henri Poincaré, Nancy I Président du jury Mme Catherine BASTIEN Chargée de Recherche, INRA Orléans Examinatrice M. Pascal FREY Chargé de Recherche, INRA Nancy Examinateur M. Jean PINON Directeur de Recherche, INRA Nancy Directeur de Thèse

UMR 1136 INRA!UHP, Interactions Arbres/Microorganismes, Pathologie Forestière Centre INRA de Nancy, Champenoux Remerciements

L'unité mixte de recherche Interactions Arbres/Microorganismes m'a accueilli durant les quatre années qu'auront duré mon travail de thèse mettant à ma disposition tous les moyens nécessaires à la réalisation des expérimentations. Une partie de ces travaux a également été effectuée en collaboration avec l'unité mixte de recherche BIOGECO de bordeaux. L'accueil chaleureux et la compétence de cette équipe ont beaucoup contribué au développement des marqueurs microsatellites, avec en particulier l'aide d'Henri Caron.

L'étude de l'agent pathogène de la rouille du peuplier a été initiée, au sein de l'équipe de pathologie forestière, par les travaux de Jean Pinon. Les connaissances et les contacts accumulés, ainsi que les multiples discussions échangées ont grandement participés à l'accomplissement de mes recherches.

La lourde tâche de mon encadrement a incombé à Pascal Frey. Je n'énumérerais pas ici l'ensemble de ses qualités scientifiques et humaines, et insisterais seulement sur les principales à mes yeux: la disponibilité, la patience, l'humour et son goût immodéré pour la chanson française (notamment Guy Béart et Bernard Lavillier).

J'adresse également mes remerciements à Cyril Dutech, Jérôme Enjalbert et François Lefèvre qui ont participé à travers la réunion de comités de thèse au développement de la réflexion sur ce sujet de recherche, et à Catherine Bastien, Michel Chalot, Tatiana Giraud et Claire Neema qui ont bien voulu évalué ce mémoire et participé à mon jury de thèse.

L'ensemble de l'équipe de pathologie forestière a réussi, chose exceptionnelle, à supporter ma présence, mon caractère et mon humour, souvent douteux, durant prêt de quatre années. Ils m'ont également soutenu et épaulé dans des domaines très variés. Je remercie donc Christine, pour le pathotypage, Béranger pour les tournées dans la Durance, Claude pour son aide dans divers domaines informatiques, Fabien qui a largement contribué à l'amélioration de ce mémoire, Benoît pour les débats contradictoires, et le courage d'Olivier qui a eu la difficile mission de réparer les ordinateurs que je m'acharnais à griller. Deux pensées spéciales pour Axelle qui m'a apporté une aide plus que précieuse tout au long de mon travail et pour Renaud, grand amateur de Speculoos et de catch, qui a supporté sans broncher la plupart de mes divagations dans ce qui nous servait de bureau.

Au delà de cette équipe, d'autres intervenants « extérieurs» m'ont aidé par leur soutien moral ou technique. Je pense notamment à François Le Tacon, Christine Delaruelle, Annegret Kohler ou Martina Peters.

Je me dois de remercier les personnes qui ont rendu plus douces les longues soirées d'hiver et plus reposant les courts week-end Nancéiens. Ceux-ci n'ont jamais refusé de boire une petite verveine en regardant un bon film ou« Très Chasse» en ma compagnie. Je parle évidemment du Clochard, d'Axel, Cat, Dav, de la Tronchette (un noble), Emilie, Foufoune, Fouf-Ia-Rage, Gaboche, Galack, Jazz, Jim, Julia, Momo, Pico-Pico, Rami, Reine, Romain, Sandrine, Séb, Taz, Tut, et d'Ln.

Enfin je tiens à rendre hommage aux deux personnes qui me supportent, dans tous les sens du terme, depuis le début, mes parents. 'jHP NANCY 1 Bibliothèque des Sciences Rue du Jardin Botanique - CS 20148 54601 VILLERS LES NANCY CEDEX SOMMAIRE

Introduction 1

1) Les peupliers naturels 1

1) Classification et biologie 1 2) Aire de répartition et écologie .3 3) Diversité génétique de P. nigra .4

II) Les peupleraies cultivées 5

1) Historique et sélection 5 2) La populiculture en France 5 3) Utilisation du bois de peuplier 6

III) Les maladies cryptogamiques du peuplier 7

IV) Le Mélèze, hôte alternant de Melampsora larici-populina 8

1) Taxonomie et aire de répartition 8 2) Sylviculture du mélèze en France et utilisation du bois 9 3) Les agresseurs du mélèze 9

V) Melampsora larici-populina, agent de la rouille foliaire .1 0

1) Taxonomie des Melampsora spp 10 2) Cycle biologique de M larici-populina 10 3) Aire de répartition de M larici-populina 12

VI) Interaction peuplier/M larici-populina 12

1) Résistance qualitative et quantitative 12 2) Aspects histologiques de l'infection du peuplier par M larici-populina 15 3) Les dégâts occasionnés par M larici-populina 16

VII) Diversité génétique des champignons phytopathogènes .16

1) Raisons de l'essor des méthodes moléculaires 17 2) Facteurs influençant la diversité génétique des champignons phytopathogènes 19 3) Connaissances acquises sur la diversité génétique de M larici-populina 22

Chapitre 1 Développement de marqueurs microsatellites chez Melampsora larici-populina 25

1) Développement de marqueurs microsatellites 25

1) Les marqueurs microsatellites 25 2) Pourquoi développer des marqueurs microsatellites chez M larici-populina? 26 3) Quelle stratégie adopter pour isoler des loci microsatellites ? •...... •...... 27 4) Caractérisation de loci microsatellites chez M larici-populina 29

II) Article publié dans Molecular Ecology Notes 30

III) Conclusion 35

Chapitre 2 : Etude de la diversité génétique à une échelle spatiale fine à l'aide d'un échantillonnage hiérarchisé et emboîté .37

Chapitre 3 : Suivi d'une épidémie dans un système en corridor et interactions entre compartiments sauvage et cultivé 69

Chapitre 4 : Effet de la migration aux échelles continentale et intercontinentale sur la structuration génétique des populations de Melampsora larici-populina 95

Conclusion et perspectives ' 123

Références bibliograhiques 127

Annexe 1 : Challenges in microsatellite isolation in fungi.. .137

Annexe 2 : Characterization of microsatellites markers in the interspecific hybrid Phytophthora alni spp. alni, and cross-amplification with related taxa ,177 Introduction Introduction

.Le peuplier est un arbre commun dans nos paysages. Des données sur les fossiles ont montré que les premiers peupliers datent au moins de la fin du paléocène (58 millions d'années, ère tertiaire) (Eckenwalder 1996). Il était déjà planté par les Romains dans les lieux publics (ce qui explique probablement l'origine de son nom). Utilisé pour marquer l'emplacement de sources en montagne ou planté en haies pour servir de brise-vent, le peuplier avait de nombreux usages. Il était de coutume dans les Pays de la Loire de planter des peupliers à la naissance d'une fille afin de constituer la dot de mariage une vingtaine d'années plus tard. Les utilisations et les représentations de cet arbre sont nombreuses et variées, et le peuplier constitue un patrimoine naturel, écologique et économique important.

1) Les peupliers naturels

1) Classification et biologie

Selon Eckenwalder (1996), le genre Populus comprend 29 espèces regroupées en six sections (Tableau 1). Ce sont des plantes ligneuses de la famille des Salicaceae. Toutes les espèces de peuplier à l'exception de Populus lasiocarpa sont dioïques, c'est-à-dire que les inflorescences mâles et femelles, qui forment des châtons, ne sont pas portées par les mêmes individus. Les peupliers ont la capacité de se reproduire par voie sexuée et asexuée. La reproduction sexuée produit de minuscules graines contenues dans une sorte de coton et disséminées essentiellement par anémochorie et hydrochorie. Celles-ci ont une viabilité réduite et germent rapidement sur sol humide, donnant naissance à des semis groupés. De nombreux cas d'hybridation interspécifique naturelle ont été décrits (Stettler et al., 1996). On peut citer par exemple l'hybridation entre Populus alba et P. tremula (P. x canescens) en Europe ou entre P. deltoides et P. trichocarpa (P. x generosa, syn. P. x interamericana) en Amérique du Nord. L'introduction de P. deltoides en Europe au XVIIIème siècle a également abouti à des hybridations naturelles avec l'espèce indigène P. nigra (P. x canadensis, syn. P. x euramericana). La reproduction asexuée est assurée par voie végétative par drageonnement à partir du système racinaire, bouturage naturel de branches et rejet à partir des souches. Ceci permet la colonisation et/ou la régénération de plants, notamment sur les berges des cours d'eau.

1 Introduction

Tableau 1 : Classification des espèces de peuplier selon Eckenwalder (1996)

Section Espèce Origine Abaso P. mexicana Wesmael Mexique Turanga P. euphratica Olivier Eurasie, Afrique P. ilicifolia Rouleau Afrique P. pruinosa Schrenk Asie Leucoides P. lasiocarpa Olivier Chine P. glauca Haines Asie P. heterophyUa L. Amérique du Nord Aigeiros P.fremontii S. Watson Amérique du Nord P. deltoides Marshall Amérique du Nord P. nigra L. Eurasie Tacamahaca P. angustifolia James Amérique du Nord P. balsamifera L. Amérique du Nord P. ciliata Royle Himalaya P. laurifolia Ledebour Eurasie P. simonii Carrière Chine P. suavolens Fisher Asie P. szechuanica Schneider Chine P. trichocarpa Torr. & Gray Amérique du Nord P.yunnanens~Dode Chine Populus P. adenopoda Maximovicz Chine P. alba L. Eurasie P. gamblei Haines Himalaya P. grandidentata Michaux Amérique du Nord P. guzmanantlensis Vasquez & Cuevas Mexique P. monticola Brandegee Mexique P. sieboldii Miquel Asie P. simaroa Rzedowski Mexique P. tremula L. Eurasie P. tremuloides Michaux Amérique du Nord

2 Introduction

2) Aire de répartition et écologie

a)

b)

c)

Figure 1 : Aire de répartition naturelle des espèces a) Populus nigra, b) P. alba et c) P. tremula (Becker et al., 1982)

3 Introduction

Toutes les espèces de peuplier (à l'exception de P. WcifoUa) ont une aire de répartition située dans l'hémisphère nord (Tableau 1). En Europe, et donc en France, trois espèces sont indigènes: le peuplier noir (P. nigra), le peuplier blanc (P. alba) et le peuplier tremble, ou tremble (P. tremula). Ces espèces ont des aires de répartition comparables qui s'étendent de la péninsule européenne jusqu'à l'ouest de la Chine (Figure 1). Toutefois, l'aire de répartition de P. tremula est plus septentrionale, comprend la Scandinavie et s'étend à toute la Russie jusqu'au Japon et au Viêt-Nam. Le peuplier noir ne se rencontre pas dans le nord de l'Europe, et même si P. nigra n'est pas présent partout, on peut le trouver sur une grande partie du pourtour méditerranéen. Enfin P. alba est la plus méridionale des trois espèces. On ne trouve le peuplier blanc que dans le sud de la France, en Alsace et en Europe centrale et orientale, ainsi que dans le nord des pays du Maghreb. Comme la plupart des peupliers, P. nigra et P. alba sont des espèces de bord de cours d'eau qui aiment les sols profonds et bien alimentés en eau, tandis que P. tremula préfère les milieux montagneux et les plaines. Ce sont trois espèces héliophiles, qui peuvent atteindre entre 25 et 35 m de hauteur pour deux mètres de diamètre à l'âge adulte et peuvent vivre entre 150 et 200 ans.

3) Diversité génétique de P. nigra

Legionnet et Lefèvre (1996) ont étudié la diversité génétique des populations de P. nigra en France à l'aide de marqueurs neutres, les isozymes. Des taux d'hétérozygotie élevés ont été trouvés, mais les différentes populations semblent assez peu différenciées entre les sites étudiés. L'hypothèse d'une recolonisation après les dernières glaciations à partir d'un nombre restreint de parents fondateurs a été émise (Legionnet 1996). Toutefois, des flux de gènes élevés pourraient également être à l'origine de ces faibles différenciations entre sites. L'aptitude à l'hybridation interspécifique met en péril la diversité des espèces indigènes de peupliers en Europe. Par exemple, les études de diversité génétique de P. nigra ont permis de montrer qu'il existe une certaine introgression des espèces hybrides ou exotiques dans les populations sauvages, par dissémination du pollen (Vanden Broek et al., 2004; Vanden Broek et al., 2005). Mais les espèces exotiques ne sont pas les seules à menacer la diversité génétique de P. nigra. Certains individus particuliers ont fait l'objet d'un bouturage de longue date et à grande échelle. C'est le cas de P. nigra var. itaUca également appelé peuplier d'Italie, reconnaissable à son port fastigé. Cette variété est plantée aussi bien à des fins ornementales, dans les parcs, qu'utilitaires (ils constituent des brise-vent efficaces lorsqu'ils

4 Introduction

sont plantés en haies et on dit que Napoléon en fit planter le long des routes pour faire de l'ombre à son armée lors de ses déplacements). Mais ce bouturage équivaut à la reproduction clonale d'un seul et même individu (qui plus est un mâle), et do~c d'un seul et unique génotype, ce qui pourrait avoir un impact important sur la diversité génétique de l'espèce. Un réseau européen (EUFORGEN) s'est ainsi formé afin de préserver les ressources génétiques de P. nigra en constituant une collection européenne d'individus sauvages et des réserves naturelles ont été mises en place en France en 1998 (Villar et al., 2004).

II) Les peupleraies cultivées

1) Historique et sélection

L'introduction d'espèces américaines au cours du XVIIlème siècle aboutit à la sélection des premiers clones de peuplier à la fin du siècle des Lumières. Ces premiers clones étaient des hybrides naturels P. x euramericana, résultant du croisement de P. nigra et de P. deltoides. Par suite, la sélection de clones fut assurée par des instituts spécialisés en Italie, aux Pays-Bas, en Belgique et en France, dès le début du XXème siècle. De nombreux critères de sélection entrent enjeu dans le choix des clones, tels que la rectitude du tronc, l'adaptation au milieu, la croissance, la qualité du bois, le comportement vis-à-vis de contraintes abiotiques et la résistance aux maladies (Villar et al., 1995). La grande majorité des cultivars utilisés en France et en Europe sont issus de croisements entre P. nigra et P. deltoides (P. x euramericana) et entre P. deltoides et P. trichocarpa (P. x interamericana). L'utilisation des différents cultivars est soumise à des effets de mode et est également très influencée par le développement de parasites (Venturia populina, Melampsora larici-populina), l'émergence de nouveaux parasites exotiques (Marssonina brunnea) ou le contournement de résistances complètes (M larici-populina).

2) La populiculture en France

Les capacités de croissance rapide ainsi que l'aptitude au bouturage ont fait du peuplier une des premières essences de feuillus en terme de récolte en France. Les peupleraies couvrent une surface d'environ 244 000 ha, ce qui représente 1,6 % de la surface forestière française, pour plus de 19 millions de m3 de volume sur pied en 2005 en France (données

5 Introduction

IFN). Les régions les plus populicoles sont le Nord-Pas-de-Calais, la Picardie, la Champagne­ Ardenne, le Centre, l'Ile-de-France et l'Aquitaine. Les peupleraies sont exploitées en futaies monoclonales équiennes à une densité variant de 156 à 278 tiges/ha selon la richesse du sol (Baméoud et al., 1982). Bien qu'il ne représente qu'une faible proportion de la surface forestière totale, le peuplier est la deuxième essence feuillue pour la production de bois d'œuvre juste après le chêne avec en moyenne 1,5 million de m3 par an récoltés en France. Ceci s'explique par la vitesse de croissance des cultivars sélectionnés qui permet une courte révolution des cultures. Cette révolution peut varier de 30 ans pour des anciens cultivars (comme P. x euramericana 'Robusta') à 15 ans pour des cultivars récents (comme P. x interamericana 'Beaupré') dans des stations favorables. Le prix moyen sur pied des peupliers est d'environ 50 euros/m3 et varie selon les cultivars et la qualité des grumes. Entre 15 et 20 % de la production est exportée, principalement vers l'Italie et l'Espagne. Il faut également noter que le peuplier peut être planté en taillis à courte rotation, comme le saule, afin de produire du bois-énergie. Toutefois ce type de plantation n'est que peu ou pas pratiqué en France. Les peupleraies françaises présentent une faible diversité clonale et sont composées quasi exclusivement d'hybrides FI P. x euramericana et P. x interamericana. Les principaux cultivars utilisés varient au cours du temps en fonction de l'émergence de nouvelles races ou de nouvelles espèces de pathogènes, mais également selon la popularité des clones. Les P. x euramericana 'Robusta', 'I 214', 'I 45/51', 'Luisa Avanzo' et 'Dorskamp', les P. x interamericana 'Beaupré' et 'Boelare' ou encore le P. trichocarpa 'Fritzi Pauley' ont fait partie des cultivars les plus plantés en France au XXème siècle.

3) Utilisation du bois de peuplier

Le bois de peuplier est un bois blanc et tendre qui se prête bien au déroulage pour la fabrication de contre-plaqué et d'emballages légers (cagettes à fruits). Il est également utilisé en sciage. Les bois sciés sont utilisés dans l'ameublement et dans le bâtiment pour la construction de charpentes. Toutefois cette dernière utilisation a été fortement concurrencée par les bois de résineux (Sales 1995). Une partie de la production est également utilisée dans la fabrication de panneaux de particules. Les plantations en taillis à courte rotation, peu répandues en France, servent à la production de pâte à papier et de bois-énergie (Barnéoud et al., 1982).

6 Introduction

Le peuplier, au delà de son intérêt économique, présente également un intérêt scientifique de premier plan. C'est en effet le premier arbre dont le génome a été entièrement séquencé. Les raisons de ce séquençage, effectué conjointement par le département de l'énergie des Etats-Unis (US-DoE, Department of Energy) et plusieurs laboratoires internationaux, sont le statut de modèle scientifique de cette plante, la faible taille de son génome (environ 480 Mb répartis sur 19 chromosomes), l'existence de protocoles de multiplication végétative et de transformation génétique et l'intérêt du peuplier en tant que puits de carbone ou source d'énergie renouvelable (Martin & Kohler, 2004). Le séquençage a été réalisé en 2003 à partir de l'ADN du clone de P. trichocarpa 'Nisqually-l' par la méthode dite en « shotgun » et la séquence partiellement annotée du génome a été rendue publique en septembre 2006 (Tuskan et al., 2006).

III) Les maladies cryptogamiques du peuplier

Les agresseurs des peupliers sont multiples et variés: des insectes (Saperdes et Chrysomelles notamment), des bactéries qui provoquent des chancres, le virus de la mosaïque du peuplier et des champignons. Parmi ces agresseurs, les maladies fongiques ont causé et causent les plus gros dégâts dans les peupleraies. Venturia populina, agent de la tavelure du peuplier, causa d'importants dégâts au début du XXème siècle, en Italie notamment. Cette maladie entraîne des défeuillaisons précoces, ce qui induit des pertes de rendement pouvant atteindre 30 % sur les cultivars sensibles. De nos jours, V populina ne provoque plus de telles pertes, des cultivars à résistance partielle ayant été sélectionnés. La brunissure des feuilles, causée par Marssonina brunnea, apparaît en Europe dans les années 1960, après que l'agent pathogène ait été introduit d'Amérique du Nord (Giorcelli & Vietto, 1998). Ce parasite a causé des dégâts dans toute l'Europe et particulièrement en Italie sur le cultivar'l 214'. Un travail de sélection a également permis de développer des cultivars à résistance partielle et de réduire ainsi les pertes occasionnées par ce champignon. Enfin les Melampsora spp., agents de la rouille foliaire, sont les parasites qui causent le plus de pertes dans les peupleraies depuis une vingtaine d'année, et notamment M larici­ populina. Les programmes de sélection ont également permis de développer des cultivars à résistance complète, mais toutes ces résistances ont été contournées à ce jour.

7 Introduction

IV) Le Mélèze, hôte alternant de M. larici-populina

1) Taxonomie et aire de répartition

Le genre Larix compte environ une quinzaine d'espèces. Le mélèze est un des rares conifères à feuillage caduc et appartient à la famille des Pinaceae. En Europe et en France, une seule espèce est présente de manière naturelle. Le mélèze d'Europe (Larix decidua) a une aire de répartition assez limitée (Figure 2). On la trouve notamment dans l'arc alpin, les massifs montagneux d'Europe Centrale comme les Carpates, les Sudètes et les Tatras et dans les plaines du Nord de la Pologne. Le mélèze d'Europe est un arbre héliophile qui aime les sols bien alimentés en eau et qui peut pousser jusqu'à une altitude de 2400 m. Pouvant atteindre 40 m de haut et un âge d'environ 500 ans, le mélèze est un des arbres emblématique des Alpes sèches.

1 r 1 1 ( \ \ . 1 ~J l " t!" ( \ \ 1

Figure 2 : Aire de répartition naturelle de Larix decidua (Becker et al., 1982)

8 Introduction

2) Sylviculture du mélèze en France et utilisation du bois

Trois espèces de mélèzes sont cultivées en France: le mélèze d'Europe (Larix decidua), le mélèze du Japon (Larix kaempferi) et des hybrides de ces deux premières espèces (L. x marschlinsii, syn. L. x eurolepis). La sylviculture du mélèze est pratiquée en futaies équiennes à une densité de 1200 tiges/ha environ. Au cours de la croissance des plants, plusieurs éclaircies sont réalisées pour atteindre des densités d'environ 130 tiges/ha à 70 ans. Le bouturage du mélèze étant difficile, la diffusion des variétés se fait principalement par voie sexuée en verger à graines. Le volume total sur pied s'élève en France à environ 22 millions de m3 en 2005 et la production totale annuelle atteint environ 670 000 m3/an, toutes espèces confondues (donnéesIF1'l). Le bois de mélèze est apprécié pour ces qualités esthétiques (notamment la couleur), mécaniques ct sa durabilité (résistance aux champignons, aux insectes, à la lumière...). C'est pourquoi il fut très tôt utilisé pour, par exemple, la construction de chalets ou de bardeaux de toiture. Le bois de mélèze peut être utilisé en extérieur sans nécessiter de traitement ou de protections spéciales pendant plusieurs dizaines d'années. Le bois de mélèze est également utilisé sous forme de parquet ou de lambris.

3) Les agresseurs du mélèze

La plupart des agresseurs du mélèze sont des insectes. La majorité de ces insectes tels que le grand scolyte (Ips cembrae), la tordeuse du mélèze (Zeiraphera diniana) et le Chermès (Adelges laricis) sont inoffensifs pour les plantations. Certains peuvent toutefois causer d'important dégâts, comme le charançon (Hylobius abiestis) ou le capricorne du mélèze (Tetropium gabrieli) sur des stations difficiles. Plusieurs maladies cryptogamiques affectent également le mélèze. L'agent pathogène le plus dommageable est sans nul doute Lachnellula willkommii, agent du chancre du mélèze. Celui-ci pose des problèmes importants dans les peuplements à basse altitude et en climat humide. C'est d'ailleurs l'une des raisons de l'utilisation du mélèze du Japon et des hybrides interspécifiques, qui présentent une bonne résistance à L. willkommii, dans la sylviculture du mélèze en Europe. Le mélèze constitue également l'hôte alternant (écidien) de plusieurs agents de rouille des peupliers et des saules tels que M larici-populina, M larici-tremulae, M larici-epitea, etc... Ces agents de rouille se développent sur les aiguilles du mélèze au printemps, mais ne causent aucune perte importante sur le mélèze.

9 Introduction

V) Melampsora larici-populina, agent de la rouille foliaire

1) Taxonomie des Melampsora spp.

Le genre Melampsora compte plus d'une cinquantaine d'espèces. Champignons basidiomycètes de la classe des Teliomycètes et de l'ordre des Urédinales (ordre qui compte plus de 7000 espèces), ils appartiennent à la famille des Melampsoraceae. Il existe au moins 13 espèces de Melampsora spp. répertoriées sur peupliers dans le monde dont 8 sont présentes en France: M larici-populina, M allii-populina, M medusae f. sp. deltoidae, M larici­ tremulae, M magnusiana, M pinitorqua, M pulcherrima et M rostrupii. Différents critères permettent de les distinguer: l'hôte écidien, la section de l'hôte télien (peuplier) et des critères morphologiques. Ces derniers concernent essentiellement les urédospores (dimensions, forme, répartition des échinulations, épaisseur de la paroi à l'équateur) et les paraphyses (dimensions, forme et épaisseur de la paroi apicale). La localisation des télies sur les feuilles de peuplier (face inférieure ou supérieure) fait également partie des clés de détermination des Melampsora spp. Trois espèces présentent une menace pour la populiculture européenne: M larici-populina, M allii-populina et M medusae f. sp. deltoidae. Les deux premières sont originaires d'Europe et la troisième d'Amérique du Nord. En Europe et en France, c'est M larici-populina qui cause le plus de dommages dans les peupleraies.

2) Cycle biologique de M. larici-populina

A l'instar de toutes les Urédinales, M larici-populina est un champignon pathogène biotrophe obligatoire. C'est un agent de rouille hétéroïque (c'est-à-dire qu'il complète son cycle de développement sur deux plantes-hôtes) et macrocyclique (c'est-à-dire que le cycle comporte cinq stades de développement et autant de types de spores) (Figure 3). Les premières infections des feuilles de peuplier ont lieu vers la fin du printemps. Ces infections primaires sont dues à des écidiospores dicaryotiques (n+n) produites par des écidies qui se développent sur les aiguilles de l'hôte écidien, le mélèze. Ces spores sont disséminées par le vent et provoquent, une dizaine de jours après l'infection de l'hôte télien (le peuplier), l'apparition d'urédies dicaryotiques (n+n) de couleur jaune orangé, sur la face inférieure des

10 Introduction feuilles. C'est à ce stade qu'a lieu la multiplication asexuée, les milliers d'urédospores produites chaque jour par chaque urédie (entre 2500 et 5000, Frey et Pinon, 2004) causant des contaminations secondaires et provoquant les symptômes observables sur la face inférieure des feuilles de peuplier (voir encadré n°1). Cette phas épidémique s'étend de fin mai à octobre, la dissémination des spores se faisant sous l'effet du vent. Durant cette période, environ une dizaine de cycles de reproduction asexuée peuvent avoir lieu. La sénescence des feuilles à ['automne provoque l'apparition de croûtes noires sur la face supérieure des feuilles de peuplier: se sont les téhes (n+n), qui sont la forme de résistance sous laquelle M larici­ populina hiverne sous nos latitudes. Au début du printemps, la caryogamie a lieu au sein des téliospores unicellulaires. Des basides c1oisonIlées (2n) se développent et produisent après méiose des basidiospores haploïdes (n), disséminées par le ve 11. Ces spores infectent les aiguilles de mélèze, donnant naissance aux spermogonies (n), quelques jours plus tard. Ces spermogonies produisent des spermaties (n) au sein de gouttelettes hyalines. C'est lors de cette phase qu'a lieu la reproduction sexuée par plasmogamie entre une spermatie (+) et un hyphe haploïde d'une spermogonie (-). La différenciation quelques jours plus tard des hyphes en de noüvelles fructificaLons, les écidies, complète le cycle de M larici-populina.

écldiospores (n+n) urédies et urédospores (n+n)

écidies (n+n)

pl m m mélèze peuplier

spermogonies et spennaties (n) télles et téllospores (n+n)

basidiospores (n~ _:+ J ami

Jl':.SideSl2nl

Figure 3 : Cycle biologique de M larici-populinu (Frey & Pinon, 2004)

Il Introduction

3) Aire de répartition de M. larici-populina

M larici-populina est une espèce Eurasienne. Son aire de répartition coïncide globalement avec celle de son hôte naturel, P. nigra. Toutefois le développement de la populiculture et la diffusion de cultivars sensibles à M larici-populina ont permis à ce parasite de se propager dans la plupart des pays qui pratiquent la culture des peupliers hybrides, y compris dans l'hémisphère Sud. Ainsi M larici-populina a été signalé dès 1918 en Amérique du Sud (Spegazzini, 1918), en Afrique du Sud en 1972 (Gibson & Waller, 1972), en Australie (Walker et al., 1974) puis en Nouvelle-Zélande (Wilkinson & Spiers, 1976), en Amérique du Nord dès 1991 (Newcombe & Chastagner, 1993) ou encore en Islande en 1999 (H. Sverrisson, communication personnelle). Les vecteurs de ces disséminations sont mal connus, même s'il semble que dans un certain nombre de cas, l'anémochorie soit responsable de la migration de M larici-populina. Un exemple d'anémochorie est donné par Wilkinson et Spiers (1976) qui supposent que la migration de M larici-populina et de M medusae d'Australie vers la Nouvelle-Zélande s'est faite via les vents trans-tasmaniens. Une dissémination par des boutures infectées n'est pas totalement écartée, même si elle reste fort hypothétique puisqu'à ce jour une survie de M larici-populina dans des bourgeons n'a jamais été démontrée (Gérard et al., 2006).

VI) Interaction peuplierlM larici-populina

1) Résistance qualitative et quantitative

M larici-populina étant un des agents pathogènes qui cause le plus de dégâts dans les peupleraies, la résistance à ce champignon fait partie des principaux critères de sélection des nouveaux cultivars. On peut distinguer deux types de résistances: la résistance qualitative (ou complète ou race-spécifique) et la résistance quantitative (ou partielle ou non race-spécifique). La résistance qualitative empêche l'infection de la plante par l'agent pathogène. Cette résistance ne s'exprime que pour une partie des génotypes de l'agent pathogène. FIor proposa la relation gène-pour-gène pour expliquer le déterminisme génétique de cette résistance qualitative (FIor, 1942; FIor 1956). L'étude de l'interaction Linum usitatissimum 1 Melampsora lini permit à FIor de démontrer par génétique mendélienne l'existence chez la plante de gènes de résistance, dominants par rapport à la sensibilité, correspondant chez

12 Introduction l'agent pathogène à des gènes d'avirulence, dominants par rapport à la virulence. Ainsi, la présence d'un allèle de résistance chez la plante et d'un allèle d'avirulence correspondant chez l'agent pathogène aboutit à l'échec de l'infection (ou interaction incompatible). Dans les trois autres combinaisons, il y a interaction compatible entre la plante et l'agent pathogène, et infection de l'hôte. La relation gène-pour-gène n'a été à ce jour démontrée, soit par génétique mendélienne, soit par clonage des gènes correspondant, que dans un nombre réduit de pathosystèmes (Barrett, 1985). Toutefois elle est soupçonnée dans de nombreuses autres interactions hôtel agent pathogène sans avoir été formellement démontrée. C'est le cas pour le pathosystème peuplier/M larici-populina. La sélection et la diffusion en Europe de cultivars de peupliers euraméricains et interaméricains à résistance complète vis-à-vis de M larici-populina ont été rapidement suivies par le contournement de ces résistances par l'agent pathogène. Ces résistances complètes sont toutes héritées du parent P. deltoides qui n'a pas co-évolué avec M larici-populina et sont donc dites « exaptées » (Newcombe 1998). A ce jour, huit virulences ont pu être identifiées (Pinon & Frey, 2005) et une gamme différentielle de clones de peuplier permet d'identifier ces virulences au laboratoire (Tableau 2).

Tableau 2 : Description de la gamme différentielle de peuplier utilisée pour le pathotypage des isolats de M larici-populina

Virulence mise en Clone Espèce évidence

'Robusta' P. x euramericana

'Ogy' P. x euramericana

'Aurora' P. xjackii 2

'Brabantica' P. x euramericana 3

'Unal' P. x interamericana 4

'Rap' P. x interamericana 5

'87B12' P. deltoides 6

'Beaupré' P. x interamericana 7

'Hoogvorst' P. x interamericana 8

13 Inb'oductiOll

Encadré oOi : SYM:PTÔMES DE LA ROUILLE DU PEUPLIER

a) b)

e) (©P.Frey, INRA)

En milieu naturel, les peupliers noirs (P. nigra) colonisent les bancs de gravier dans le lit majeur des rivières. Les photos a) et b) montrent l'exemple de la vallée de la Durance où l'on peut distinguer les cohortes de peupliers noirs d'âges ditIérents (photo a) et les brosses de semis (photo b). L'infection des peupliers noirs par M larici-populina se traduit par la présence de pustules jaune-orangé, les urédies, sur la face inférieure des feuilles. La photo c) montre des feuilles de peuplier noir sauvage infectées par M larici-populina dans la ripisylve de la Durance.

Les peupleraies cultivées sont la plupart du temps organisées en futaies monoclonales équiennes. Lorsqu'une parcelle est fortement atteinte par la rouille du peuplier, la couleur des feuilles vire à la couleur jatme-orangé, voire brune, dès la fin de l'été. C'est ce que j'on peut observer sur la photo d) qui montre une plantation de peupliers interaméricains (P. x interamericana 'Beaupré') fortement infectée par M. larici­ populina dans la vallée du Rhône. De même que pour les P. nigra, les feuilles de peuplier euraméricmn présentent sur leur face intërieure des urédies jaune-orangé (feuilles de P. x euramericana '145-51' naturellement infectées, photo e)).

14 Introduction

La stratégie de sélection de peupliers hybrides euraméricains et interaméricains possédant des résistances qualitatives s'est donc révélée inefficace dans la lutte contre M larici­ populina. La faible diversité des cultivars employés ainsi que la pérennité des plantations de peupliers sont autant de facteurs expliquant les contournements rapides et successifs des résistances advenus ces dernières années. Ce sont également des facteurs qui peuvent aggraver l'impact des épidémies de rouilles. Ainsi le contournement du gène de résistance R7 en 1994 a provoqué dans les années suivantes de graves épidémies sur des cultivars extrêmement répandus, tels que 'Beaupré' et 'Boelare', et qui possédaient une faible résistance quantitative à la rouille. La résistance quantitative est une résistance non race-spécifique, c'est-à-dire qu'elle s'exerce à priori sans distinction sur l'ensemble des génotypes de l'agent pathogène. Contrairement à la résistance qualitative, elle n'empêche pas l'infection mais retarde ou limite les symptômes de la maladie. Cette résistance de nature polygénique est supposée être plus durable, mais la sélection de tels caractères est complexe et coûteuse, notamment chez le peuplier. La résistance partielle est principalement héritée des parents P. trichocarpa et P. nigra (Bastien et al., 2004). Un premier gène ou groupement de gènes majeurs de résistance quantitative a été identifié dans une famille FI P. deltoides x P. trichocarpa (Dowkiw et al., 2003 ; Jorge et al., 2005). Le développement de résistance durable à M larici-populina passera sans doute par de nouvelles constructions génétiques (par exemple le pyramidage des gènes de résistance) et par le développement de nouvelles résistances complètes et partielles. Le déploiement raisonné des différents cultivars dans le temps et l'espace ou la pratique du mélange variétal peuvent également être envisagés. Toutefois une première étude sur le mélange variétal de clones de peuplier n'a pas pu mettre en évidence d'effets positifs significatifs sur la croissance (Miot et al., 1999).

2) Aspects histologiques de l'infection du peuplier par M. larici-populina

Après le dépôt d'urédospores sur la surface inférieure de la feuille de peuplier, celles-ci produisent des tubes germinatifs. Les conditions de température et surtout d'humidité jouent un rôle important lors de cette phase. Ainsi la mouillabilité des feuilles, qui influe sur la vitesse d'évaporation de l'eau à la surface de la feuille, joue un rôle sur le succès de l'infection par les spores (Pinon et al., 2006). Lorsqu'un tube germinatif arrive à proximité d'un stomate, il se différencie en appressorium. Celui-ci produit un tube infectieux qui

15 Introduction

pénètre dans la chambre sous-stomatique et forme une vésicule (Spiers & Hopcroft, 1988). A partir de cette vésicule, des hyphes infectieux septés croissent et s'insinuent entre les cellules des parenchymes lacuneux et palissadiques (Laurans, 1997). Au contact des cellules du parenchyme, les hyphes forment des haustoria, qui sont des structures spécialisées dans les échanges trophiques entre la plante et le champignon. En conditions optimales, il s'écoule environ 17 heures entre l'inoculation de M larici-populina et la formation des premiers haustoria (Laurans, 1997). Dans le cas d'une interaction incompatible, on observe alors une réaction d'hypersensibilité localisée: les cellules végétales envahies par les haustoria subissent une plasmolyse de leur cytoplasme et les hyphes fongiques sont détruits. Lorsque l'interaction est compatible, M larici-populina se développe en exploitant les ressources prélevées sur son hôte. Environ sept jours après l'inoculation, l'agrégation des hyphes intercellulaires sous l'épiderme donnera naissance à des urédies (Laurans & Pilate, 1999).

3) Les dégâts occasionnés par M. larici-populina

M larici-populina est un champignon biotrophe obligatoire qUl pUlse ses ressources carbonées et azotées sur son hôte. L'infection des feuilles entraîne une diminution de la photosynthèse et une augmentation de la respiration (Chiba & Kohda, 1985). Les perturbations engendrées par l'agent pathogène entraînent des défeuillaisons précoces au cours de l'été. Ces défeuillaisons partielles, voire totales lorsque l'infection est sévère, affectent les mises en réserves qui ont normalement lieu au début de l'automne. De plus, l'aoûtement des bourgeons peut être perturbé, aboutissant même parfois à des débourrements prématurés. Des pertes de croissance importantes ont pu être constatées: jusqu'à 60 % de perte de croissance en circonférence après plusieurs épidémies répétées de rouille (Gastine et al., 2003). L'affaiblissement des arbres qui résulte de ces infections peut entraîner leur mort s'ils subissent en plus l'attaque de parasites secondaires, tel que Discosporium populeum ou Cytospora chrysosperma (Wang & Van der Kamp, 1992).

VII) Diversité génétique des champignons phytopathogènes

La diversité génétique des champignons phytopathogènes a été étudiée depuis de nombreuses années à l'aide de caractères phénotypiques et notamment par l'étude de leur pouvoir pathogène. Les premières études démontrant la relation gène-pour-gène dans le

16 Introduction

pathosystème Linum usitatissimum/Melampsora Uni ont par exemple été réalisées dès le début des années 1940 (FIor 1942). Toutefois depuis une vingtaine d'années, l'étude de la diversité génétique à l'aide de marqueurs moléculaires c'est considérablement développée (Luikart & England, 1999).

1) Raisons de l'essor des méthodes moléculaires

Les premiers marqueurs moléculaires utilisés pour les études de diversité génétique furent les isoenzymes (ou isozymes) au début des années 1970. Les progrès constants des connaissances en biologie moléculaire et en particulier le développement de la réaction d'amplification en chaîne (PCR, Polymerase Chain Reaction; Mullis & Faloona, 1987) ont permis le développement de nombreux types de marqueurs moléculaires dans les années 1990, ayant chacun des caractéristiques propres (voir encadré n02). De manière générale, l'avènement des marqueurs moléculaires a permis l'accès à un niveau de polymorphisme fin et à une exploration plus systématique du génome. La multiplication du nombre de loci (avec notamment les marqueurs révélés en masse comme les RAPD ou les AFLP) ou l'augmentation du nombre d'allèles mis en évidence par locus (notamment chez les marqueurs microsatellites) ont entraîné un accroissement important de la qualité de l'information collectée. Associé à ces progrès qualitatifs des marqueurs, des développements importants dans la miniaturisation et l'automatisation ont abouti à des méthodes d'analyse du polymorphisme dites de haut débit. Il est aujourd'hui possible de génotyper rapidement des populations de plusieurs centaines d'individus sur de nombreux loci, ce qui est tout à fait adapté à la réalisation d'études de génétique des populations. La quantité de données ainsi générées serait difficilement exploitable si de manière concomitante il n'y avait eu un développement de l'informatique et des méthodes analytiques. Ces dernières années, de nombreux logiciels de génétique des populations ont été développés et sont disponibles en accès libre sur internet. Ils permettent l'exploitation des jeux de données obtenues grâce aux nouvelles techniques de biologie moléculaire.

Tous ces facteurs ont abouti à un renouveau des études de génétique des populations, y compris chez les champignons phytopathogènes.

17 Introduction

Encadré n02 : LES PRINCIPAUX MARQUEURS MOLECULAIRES

Les isoenzymes : le principe de ces marqueurs repose sur le polymorphisme de séquences codant certaines enzymes. Ces marqueurs sont le plus souvent codominants, mais leur développement et leur utilisation peuvent se révéler techniquement complexe. Le caractère neutre de ces marqueurs a également été remis en cause. Ils furent par exemple utilisés dès 1983 pour étudier le polymorphisme de différentes espèces de Puccinia (Burdon et al., 1983).

Les RFLP (Restriction Fragment Length Polymorphism) mettent en évidence un polymorphisme de séquences au niveau des sites de restriction. Contrairement aux isoenzymes, ce polymorphisme est révélé directement au niveau de la séquence d'ADN. Marqueurs codominants, les RFLP nécessitent des quantités d'ADN assez importantes. Une des premières utilisations de ces marqueurs chez un champignon phytopathogène fut l'étude de McDonald et Martinez (1990) sur Mycosphaerella graminicola.

Les RAPD (Random Amplified Polymorphie DNA) sont des marqueurs dominants. Leur principe repose sur une amplification de l'ADN à l'aide d'oligonucléotides de faible taille (en général dix nucléotides). Cette méthode permet de mettre en évidence beaucoup de polymorphisme réparti aléatoirement dans le génome. Cette technique relativement aisée à développer, présente l'inconvénient majeur d'être difficilement transférable entre laboratoires. De nombreuses études ont été réalisées à J'aide de RAPD sur des champignons phytopathogènes, parmi lesquelles des études sur Fusarium solani f. sp. cucurbitae (Crowhurst et al., 1991) ou Leptosphaeria maculans (Goodwin & Annis, 1991).

Les AFLP (Amplified Fragment Length Polymorphism) consistent en l'amplification sélective de fragments de restriction d'ADN génomique. Ce sont des marqueurs dominants. Ils ont entre autre été utilisés pour des études de diversité génétique de Melampsora larici­ epitea (Samils et al., 2001) et de Melampsora larici-populina (Pei et al., 2005).

Les microsatellites : (ou SSR pour Simple Sequence Repeat) sont des répétitions en tandem de motifs d'ADN d'une à six paires de base. Ce sont des marqueurs codominants dont le polymorphisme repose sur la variation du nombre de répétitions du motif nucléotidique. Parmi les premiers exemples d'utilisation de tels marqueurs chez les champignons on peut citer Aspergillus fumigatus (Bart-Delabesse et al., 1998) ou encore Mycosphaerella fTjiensis (Neu et al., 1999).

Les SNP (Single Nucleotide Polymorphism) sont des marqueurs génétiques qui mettent en évidence le polymorphisme au niveau d'un seul nucléotide. Leur développement nécessite la possession d'un nombre important de données génomiques. Leur utilisation, certes coûteuse, semble toutefois prometteuse pour des études de diversité génétique.

18 Introduction

2) Facteurs influençant la diversité génétique des champignons phytopathogènes

Comme pour tous les autres organismes vivants, les mutations de l'ADN sont à l'origine de la diversité génétique observée chez les champignons phytopathogènes. Il a été montré par exemple que le contournement d'une résistance chez la tomate par Cladosporium fulvum résulte de la mutation d'un seul nucléotide (Joosten et al., 1994). Toutefois l'importance et l'organisation de cette diversité génétique créée par des mécanismes mutationnels résulte de l'interaction de nombreux autres facteurs.

a) La recombinaison

Une des particularités du règne fongique est la diversité des modes de reproduction qui vont de l'absence de sexualité à des étapes de reproduction sexuée obligatoire, en passant par des phénomènes de parasexualité (Brygoo et al., 1998). La reproduction sexuée permet la création de nouveaux génotypes via la recombinaison de gènes. Chez les champignons phytopathogènes, cela conduit par exemple à l'association d'allèles de virulence dans différents fonds génétiques, ce qui rend potentiellement caduque une stratégie de pyramidage de gènes de résistance qualitative chez l'hôte (McDonald 1997). Il a été montré que l'absence de reproduction sexuée chez Puccinia graminis f. sp. tritici, agent de la rouille noire du blé, suite à l'éradication de son hôte alternant, l'épine vinette, des grandes plaines céréalières des Etats-Unis, entraînait une augmentation des déséquilibres de liaison et une diminution de la diversité génétique dans les populations étudiées (Roelfs & Groth, 1980; Groth & Roelfs, 1982). On peut également citer les études réalisées à l'aide de marqueurs isoenzymatiques sur l'agent du mildiou de la pomme de terre, Phytophthora infestans. La reproduction sexuée chez cet Oomycète est conditionnée par la présence de deux types sexuels Al et A2. Le polymorphisme des marqueurs génétiques s'est révélé plus élevé dans les populations possédant les deux types sexuels (Tooley et al., 1985 ; Sujkowski et al., 1994)

b) La sélection par l'hôte

Les effets de la pression de sélection exercée par l'hôte sur la diversité génétique des agents phytopathogènes ont été notamment observés dans les systèmes agricoles mais ils ont également été montrés dans des pathosystèmes naturels.

19 Introduction

La sélection, puis le déploiement à grande échelle de variétés résistantes dans les grandes cultures, a souvent entraîné l'apparition puis le développement de pathotypes virulents qui ont fortement modelé la structure des populations. Il a ainsi été observé une divergence des populations de Puccinia recondita f. sp. trUici en Amérique du Nord après l'introduction de cultivars de blé résistants au Canada (Kolmer 1991a; Kolmer 1991b). Bien que l'influence d'une pression de sélection exercée par l'hôte ait des effets importants sur la fréquence des marqueurs sélectionnés, elle peut également entraîner une différenciation génétique neutre résultant d'un effet de fondation. Ainsi le contournement récent d'un gène de résistance du pommier par Venturia inaequalis a provoqué une forte différenciation et une baisse de la diversité génétique chez les populations virulentes (Guérin & Le Cam, 2004). Dans les pathosystèmes naturels, la sélection entre l'hôte et l'agent pathogène est réciproque, ce qui entraînerait le polymorphisme des gènes de résistance et d'avirulence (Thrall & Burdon, 2003). Des patrons d'adaptation locale d'agents pathogènes à leurs hôtes ont été mis en évidence dans les interactions entre la légumineuse Amphicarpea bracteata et le champignon Synchytrium decipiens (Parker 1985) ou encore entre le haricot (Phaseolus vulgaris) et l'agent de l'anthracnose (Colletotrichum lindemuthianum) (Geffroy et al., 1999). Dans le cas des rouilles, l'adaptation locale de l'agent pathogène à son hôte a par exemple été montrée dans les pathosystèmes Linum marginale/Melampsora lini (Thrall & Burdon, 2003) ou Salix triandra/Melampsora amygdalinae (Niemi et al., 2006). La variabilité des patrons de virulence et de résistance serait la conséquence de phénomènes d'extinction-recolonisation et de dérive locale (Burdon 1992). Des simulations ont montré qu'une structure en métapopulations et l'influence des distances de dispersion pourraient théoriquement expliquer le maintien des niveaux de polymorphisme observés sans assumer un coût des virulences (Thrall & Burdon, 2002).

c) Les flux de gènes

La dissémination des champignons phytopathogènes s'effectue le plus souvent de manière passive. Les vecteurs de dispersion, ainsi que les distances parcourues, sont très variables. On peut citer par exemple le vent, l'eau, les insectes ou l'homme comme vecteurs de dissémination. La dissémination d'agents pathogènes peut aboutir à la colonisation de nouvelles zones géographiques, ce qui peut avoir des effets catastrophiques sur les agro­ écosystèmes ou sur les écosystèmes naturels. Un exemple historique est l'introduction de Phytophthora infestans en Europe en 1845 via des tubercules infectés, qui entraîna des dégâts

20 Introduction

considérables sur les cultures de pomme de terre et la terrible famine irlandaise (Fry et al., 1992). L'augmentation des échanges internationaux au cours du XXe siècle a provoqué la dispersion globale de nombreux agents pathogènes. Certains champignons ont ainsi pu coloniser la quasi-totalité des régions où leur hôte était présent, comme Mycosphaerella jijensis, l'agent de la maladie des raies noires du bananier (Rivas et al., 2004). La migration est définie comme le mouvement d'organismes entre des sous-populations. Elle tend à l'homogénéisation des sous-populations. Les études à l'aide de marqueurs neutres ont permis ces dernières années d'évaluer l'importance de la migration dans certains pathosystèmes. Dans le cas de l'agent de la septoriose du blé, Mycosphaerella graminicola, il a été montré que la différenciation entre les populations à l'échelle mondiale était assez faible, impliquant des flux de gènes relativement importants (Linde et al., 2002). En revanche, dans le cas de la maladie des raies noires du bananier, les différentes populations de M fzjensis étaient bien plus différenciées à l'échelle intercontinentale et même intracontinentale (Rivas et al., 2004). Toutefois ces estimations résultent de la combinaison des capacités de dispersion et des traits d'histoire de vie des champignons. Ainsi les échanges de blé, potentiellement vecteurs de M graminicola, sont très anciens alors que la culture du bananier au niveau mondial, et donc la dissémination de M fijensis, sont bien plus récentes. La plupart des Urédinales produisent d'énormes quantités d'urédospores en phase épidémique (Robert et al., 2002). Ces spores ont la capacité de se disperser sur de très longues distances et de nombreux cas de migrations intercontinentales ont été observés (pour une synthèse voir Nagarajan & Singh, 1990). Une multiplication asexuée importante, combinée à une dispersion à longue distance peut aboutir à une structure clonale des populations sur des étendues très grandes, comme cela a été observé pour les rouilles des céréales en Amérique du Nord (Leonard et al., 1992; Long et al., 1993) ou en Europe du Nord (Hovm011er et al., 2002).

La migration est considérée comme une force agissant contre la divergence génétique entre sous populations. En cela, elle s'oppose à la dérive génétique.

d) La dérive génétique

A chaque génération, on peut considérer qu'un tirage aléatoire de gamètes a lieu pour former la génération suivante. Cette sélection au hasard peut affecter les fréquences allèliques, aboutissant à la fixation ou à la disparition de certains allèles. Il en résulte des variations de la

sen UHP NANCY 1 Bibliothèqùe des Sciences 21 Rue du Jardin Bot.anique- CS 20148 54601 VILLERS LES NANCY CEDEX Introduction

diversité génétique. On nomme ce procédé la dérive génétique. L'importance de la dérive est principalement liée à la taille de la population. Plus celle-ci est faible, plus l'effet de la dérive sera important. Le cycle des champignons phytopathogènes comprend souvent une phase de forte diminution de la taille de la population. En particulier, pour les champignons biotrophes obligatoires, la disponibilité en tissus végétaux colonisables est variable selon les saisons en climat tempéré, ce qui peut entraîner des goulots d'étranglement des populations pathogènes (en hiver sur les plantes à feuilles caduques et en été sur les céréales). De telles fluctuations de la taille des populations pathogènes pourraient donc engendrer une forte dérive génétique. Cependant, une étude sur trois années successives d'une population de Mycosphaerella graminicola à l'aide de marqueurs neutres n'a pas mis en évidence d'effet majeur de la dérive sur la diversité de cet agent pathogène (Chen et al., 1994).

3) Connaissances acquises sur la diversité génétique de M. larici-populina

Les premiers marqueurs utilisés pour étudier la diversité génétique de M larici-populina furent des marqueurs sélectionnés. Ces facteurs de virulence sont aujourd'hui au nombre de huit et ont permis de mettre en évidence une forte différenciation entre ies populations des compartiments sauvage et cultivé (Pinon & Frey, 1997). Les populations de M larici­ populina prélevées dans le compartiment sauvage (ripisylves à Populus nigra) présentent des pathotypes simples, avec une prédominance du pathotype '0' (ne possédant aucune virulence), au contraire des individus issus du compartiment cultivé (plantations de peupliers hybrides) qui possèdent des pathotypes complexes et variés. Les pressions de sélection successives exercées par le déploiement de cultivars de peupliers hybrides complètement résistants à M larici-populina semblent avoir entraîné un pyramidage des facteurs de virulence, ayant pour conséquence une complexification grandissante des pathotypes (Pinon & Frey, 2005). De plus, un effet significatif de la présence du mélèze (l'hôte alternant) sur la richesse pathotypique a été mis en évidence (Frey et al., 2005). Parallèlement à ces études, des marqueurs de type RAPD ont été développés dans le but d'étudier la diversité neutre des populations de M larici-populina (Foulon 1999). Ceux-ci ont permis de mettre en évidence une très grande diversité génétique aussi bien au sein des populations échantillonnées dans le compartiment sauvage que dans le compartiment cultivé. La différenciation modérée entre les différentes populations, ainsi que les faibles déséquilibres de liaisons mesurés entre les marqueurs RAPD dans les populations éloignées de l'hôte alternant, sont des indicateurs d'une dispersion à grande échelle (Gérard et al.,

22 Introduction

2006). Aucun patron d'isolement par la distance n'a pu être mis en évidence lors de cette étude, et aucune corrélation entre les structures pathotypiques et neutre n'a été révélée. Même si ces études nous offrent de précieux renseignements sur l'organisation de la diversité génétique de M larici-populina, il est important de souligner que l'utilisation de marqueurs dominants pour l'analyse d'individus dicaryotiques présente des inconvénients.

Dans l'optique de la réalisation d'études fines de la diversité génétique neutre de M larici-populina, qui pourront potentiellement apporter des connaissances précieuses sur la biologie et l'épidémiologie de ce champignon pathogène, l'objectif de cette thèse a tout d'abord été le développement de marqueurs génétiques neutres mieux adaptés aux études de génétique des populations d'un organisme dicaryotique. Le choix s'est porté sur des marqueurs codominants et réputés très polymorphes: les microsatellites. Suite à l'obtention de ces marqueurs et afin de mieux appréhender l'organisation fine de la diversité génétique de M larici-populina, l'étude de trois populations françaises échantillonnées de manière hiérarchisée et emboîtée sur quatre niveaux (feuille, rameau, arbre et site) a été entreprise. Par aiileurs, l'effet de la progression d'une épidémie de rouiile à partir d'une zone de sympatrie peuplier-mélèze vers des zones où seul le peuplier est présent, sur la diversité génétique a été étudié sur des populations de la ripisylve de la vallée de la Durance. Les éventuels échanges entre les populations du compartiment cultivé et du compartiment sauvage ont également été suivis dans ce système en corridor. Enfin, la dissémination par le vent des agents pathogènes tel que M larici-populina, permet souvent une migration à grande distance des spores et d'importants flux géniques. Ces phénomènes sont extrêmement complexes à étudier de manière directe. C'est pourquoi la diversité de huit populations du continent européen ainsi que de deux populations récemment fondées en Islande et au Canada a été étudiée. Le but de cette dernière étude était de mettre en évidence les effets de l'éloignement géographique sur la différenciation des populations de M larici-populina.

23 Introduction

24 Chapitre 1

Le développement de marqueurs microsatellites chez Melampsora larici-populina Développement de marqueurs microsatellites

1) Développement de marqueurs microsatellites

1) Les marqueurs microsatellites

Les marqueurs microsatellites sont définis comme des répétitions en tandem de motifs nucléotidiques de 1 à 6 paires de bases. Egalement appelés SSR (Simple Sequence Repeat), VNTR (Variable Number of Tandem Repeat) ou STR (Short Tandem Repeat), ils ont été découverts dans la plupart des taxons eucaryotes (insectes, poissons, reptiles, mammifères, angiospermes, etc...) (Li et al., 2002 ; Ellegren, 2004). L'accès récent à la séquence complète de génomes a permis de démontrer que les loci microsatellites étaient largement présents et dispersés au sein des génomes, préférentiellement dans des parties non codantes (Hancock, 1995). Toutefois, des loci microsatellites ont aussi été identifiés dans des régions codantes des génomes. Les motifs les plus fréquents sont alors ceux qui entrent dans le cadre de lecture du code génétique (i.e. des motifs trinucléotidiques ou héxanucléotidiques, T6th et al., 2000 ; Metzgar et al., 2000). L'organisation de la chromatine, la régulation de gènes ou un rôle lors de la recombinaison, sont autant de fonctions imputées aux microsatellites (Kashi & Soller, 1999).

Ces marqueurs possèdent trois caractéristiques importantes, ils sont codominants, réputés très polymorphes et souvent neutres (Jarne & Lagoda, 1996) :

CD La codominance permet, dans le cas d'organismes diploïdes, de distinguer les produits de deux allèles distincts. Ceci augmente la capacité résolutive des marqueurs et permet le calcul de paramètres génétiques comme l'hétérozygotie ou le coefficient de consanguinité. Les isoenzymes, comme les microsatellites, sont des marqueurs codominants, au contraire des marqueurs RAPD ou des AFLP, qui sont dominants.

CD Un locus est dit polymorphe lorsque son allèle le plus répandu a une fréquence inférieure à 95 % (Hartl & Clark, 1997). Les microsatellites sont généralement des loci très polymorphes, c'est à dire qu'ils possèdent de nombreux allèles (parfois plusieurs dizaines). Ce polymorphisme est la conséquence d'un fort taux de mutation (compris entre 2 6 10- et 10- , Luikart & England, 1999; Xu et al., 2000), lui même étant probablement la conséquence de phénomènes de mésappariements lors de la réplication de l'ADN (Levinson & Gutman, 1987). Ceci aboutit à une variation du nombre de répétitions du motif microsatellite (Eisen, 1999). Le polymorphisme des microsatellites est donc un

25 Chapitre 1

polymorphisme de nombre de répétitions, qUI se traduit concrètement par un polymorphisme de taille du fragment amplifié.

e Un autre aspect intéressant des marqueurs microsatellites est leur neutralité. On dit d'un locus qu'il est neutre lorsqu'il n'est pas soumis à des pressions de sélection, mais seulement à la dérive génétique et aux flux de gènes entre populations. Cela en fait donc des loci adaptés à des études de génétique des populations. Toutefois, il faut noter qu'il existe des contre-exemples à cette neutralité. Il a été démontré par exemple que certaines maladies génétiques humaines, comme la Chorée de Hungtington, sont directement causées par la mutation d'un locus microsatellite (Ranum & Day, 2002).

Toutes les qualités de ces marqueurs les rendent particulièrement précieux et font qu'ils sont couramment utilisés dans des domaines aussi variés que la détermination de la paternité, l'identification par empreinte génétique, la réalisation de cartographies génétiques ou les études populationnelles (Luikart et al., 2003 ; Sunnucks, 2000; Selkoe & Toonen, 2006).

2) Pourquoi développer des marqueurs microsatellites chez M. larici-populina ?

Outre les qualités des microsatellites en tant que marqueurs génétiques, il existe deux autres caractéristiques chez ces marqueurs qui sont intéressantes pour l'étude d'un champignon pathogène tel que M larici-populina. La première est que les microsatellites sont des marqueurs très sensibles, qui reposent sur le principe de l'amplification de fragments d'ADN. Ils requièrent donc une faible quantité d'ADN, ce qui est un avantage lorsque l'on étudie un microorganisme biotrophe obligatoire dont il est difficile d'obtenir de l'ADN en grande quantité. Le caractère biotrophe du champignon pathogène nous amène à considérer le deuxième aspect intéressant des loci microsatellites pour l'étude de M larici-populina: leur spécificité. L'amplification des SSR se fait avec un couple d'amorces spécifiques. L'échantillon d'ADN du champignon peut donc être extrait en mélange avec l'ADN d'un autre organisme sans que cela affecte l'analyse. Ainsi dans le cas de M larici-populina, on peut envisager de réaliser des analyses à partir d'ADN extrait directement sur des fragments de feuille de peuplier portant une urédie, comme cela a déjà été décrit pour M medusae f. sp. deltoidae (Bourassa et al., 2005).

26 Développement de marqueurs microsatellites

3) Quelle stratégie adopter pour isoler des loci microsatellites ?

Les marqueurs microsatellites sont donc particulièrement bien adaptés aux études de génétique des populations sur M larici-populina. Ils ont toutefois un inconvénient majeur: ils doivent être développés de nova pour chaque nouvelle espèce étudiée. Même si leur nombre absolu dans un génome eucaryote peut paraître important (environ 16000 loci identifiés dans le génome de Magnaporthe grisea, Li et al., 2004), il peut malgré tout s'avérer difficile d'isoler des loci d'intérêt. Cette difficulté semble même accrue dans le groupe des champignons. En effet, à ce jour, des loci SSR n'ont été développés que dans une cinquantaine d'espèces fongiques, avec plus ou moins de succès (Dutech et al., soumis). L'étude de l'abondance des loci microsatellites dans des génomes de champignons a montré qu'il y avait moins de loci et avec moins de répétitions chez les champignons que dans d'autres groupes phylogénétiques (T6th et al., 2000 ; Lim et al., 2004, Karaoglu et al., 2005). Il semble également que ces loci soient plus difficiles à isoler et moins polymorphes (Dutech et al., soumis). Pour pallier ces difficultés, il existe plusieurs stratégies de développement de marqueurs microsatellites. Tout d'abord, si on a la chance de connaître la séquence génomique complète de l'organisme étudié ou si d'importantes banques de séquences de fragments exprimés (EST, Expressed Sequence Tag) ont été développées préalablement, on peut réaliser une recherche in silico de motifs microsatellites. Il existe plusieurs logiciels en accès libre pour réaliser ce type de recherche, comme Magellan (Lim et al., 2004). Par ailleurs, lorsque des marqueurs ont été décrits dans une ou des espèces proches, il est possible de tenter de transférer ces loci sur l'espèce que l'on étudie. L'efficacité de cette méthode dépend notamment de la distance génétique qui sépare les deux espèces et peut parfois s'avérer payante, même si elle peut également entraîner une réduction du polymorphisme des loci transférés (Petit et al., 2005). Enfin, une des méthodes les plus couramment utilisées pour isoler des loci microsatellites est la réalisation de banques enrichies en motifs microsatellites (pour une revue des méthodes voir Zane et al., 2002). Au début de ma thèse, l'absence de données génomiques et l'inexistence de marqueurs microsatellites décrits dans une espèce phylogénétiquement proche de M larici-populina nous a conduit à choisir cette dernière méthode. Il faut également noter que cinq autres loci microsatellites ont été développés entre­ temps pour M medusae f. sp. deltoidae et M larici-populina (Steimel et al., 2005).

27 Chapitre 1

Tableau 3 : Résultats des différentes banques enrichies en microsatellites de Melampsora larici-populina

Enrichissement 2 3 4 5 Total

Méthode Membranes Membranes Billes Billes Billes d'enrichissement

Motifdes sondes (AC) 15 et (GAC)10 (AC)15 (TC) 10 (TG) 10 utilisées (AG)15 Nombre de clones 192 96 90 1920 1920 4218 bactériens isolés Amplification de Hybridation d'une Hybridation d'une Type de criblage aucun aucun l'ADN (PCR) sonde (Colony Blot) sonde (Colony Blot) Proportion de clones 20% 15% 35% positifs au criblage Nombre de séquences 104 8 16 40 40 208 réalisées

Séquences redondantes 31 3 4 5 2 45

Nombre de loci 20 0 2 9 6 37 identifiés

Taux d'enrichissement 0,19 0,00 0,13 0,23 0,15 0,18

Nombre de paires 28 0 3 14 7 52 d'amorces définies Nombre de loci 3 0 2 5 5 15 exploitables Nombre de loci 3 0 2 4 4 13 polymorphes

28 Développement de marqueurs microsatellites

4) Caractérisation de loci microsateUites chez M. larici-populina

L'isolement des loci microsatellites chez M larici-populina a fait l'objet d'une publication dans Molecular Ecology Notes (voir ci-après). La difficulté à isoler des loci d'intérêt chez cet agent pathogène nous a conduits à réaliser cinq banques enrichies à l'aide de deux techniques d'enrichissement différentes, sur membrane de nylon et sur billes magnétiques (Tableau 3). Une étape de criblage supplémentaire dans trois de ces banques enrichies a été réalisée. Deux protocoles de criblage ont été utilisés. Le premier criblage a été effectué par amplification de l'ADN en utilisant une amorce [TG]g ou [TC]s en combinaison avec une amorce du vecteur de clonage (ici pCR 4-TOPO). Ainsi, en principe, lorsqu'un locus microsatellite est présent dans l'insert du vecteur de clonage, il y a une amplification. Pour augmenter la spécificité de la méthode, un gradient de température a été testé pour chaque couple d'amorces afin de déterminer la température d'hybridation idéale. Malheureusement, il est apparu que les amorces [TG]s ou [TC]s pouvaient se fixer sur des loci à faible nombre de répétitions (3 ou 4). Or les loci microsatellites avec un nombre de répétitions aussi faible sont en général monomorphes et ne sont donc pas utilisables pour des analyses de génétique des populations. C'est pourquoi nous avons abandonné cette technique. La seconde méthode de criblage est inspirée directement du protocole décrit par Estoup et al. (1993). Nous avons ainsi criblé les deux banques de 1920 clones, qui avaient été enrichies à l'aide de billes magnétiques. Quelques modifications mineures ont été apportées au protocole original de criblage, portant principalement sur les volumes des solutions (Guérin et al., 2004). Les colonies bactériennes ont été déposées par simple contact sur des membranes de nitrocellulose chargée positivement (Hybond-N+, Amersham) après une culture sur un milieu LB agar additionné d'ampicilline. L'hybridation (ou «colony blot») est réalisée avec des sondes

[TG]1O ou [TC] 10 marquées à la digoxygénine, et la révélation est effectuée grâce à un kit de détection par colorimétrie (Roche). Au total, à partir de cinq banques enrichies, plus de 4200 clones ont été isolés, environ 980 clones se sont révélés positifs après l'étape de criblage, puis les inserts de 208 clones ont été séquencés pour aboutir à l'isolement de 15 loci microsatellites dont 13 polymorphes (Tableau 3). Ceci illustre l'important investissement en temps pour isoler de tels marqueurs chez un champignon. Ces difficultés nous ont d'ailleurs conduits à participer à la rédaction d'un article de synthèse sur la difficulté de développer des marqueurs microsatellites chez les champignons (Dutech et al., soumis, voir Annexe 1).

29 Chapitre 1

II) Article publié dans Molecular Ecology Notes

Molecular Ecology Notes (2006) 6, 60~64 doi: 10.1111/j.1471-8286.2005.01137.x

PRIMER NOTE Isolation and characterization of 15 microsatellite loci in the poplar rust fungus, Melampsora larici-populina, and cross-amplification in related species

BENOîT BARRÈS,* CYRIL DUTECH,tAXELLE ANDRIEUX,*HENRI CARON,t JEAN PINON* and PASCAL FREY* *UMR IAM, PathnlogieForestière, INRA Nancy, 54280 Champenoux, France, tUMR BIOGECO, INRA Pierroton, 33612 Cestas, France

Abstract We developed 15 microsatellite loci in the poplar rust fungus,Melampsora larici-populina, using two enrichmentprotocols.Polymurphism ofeach locus was assessed on a panel of30 isolates, comprising three subpanels (world, regional and local scales). Thirteen loci were polymorphie with three toeightalleles detected. The 15 loci were also tested on five related Melampsora species, M. allii-populina, M. medusae f. sp. deltoidae, M. larici-tremulae, M. rostrupii and M. pinitorqua, and partial or global cross-amplification events were detected. Keywords: Basidiomycete, cross-species amplification, enriched library, Melampsora larici-populina, microsatellites Received 24 June 2005; revision received 8 July 2005; aeeepted 18 Iuly 2005

The Basidiomycete Melampsora larici-populina causes enriched library for microsatellites. Total genomic DNA follar !TIst on Populus species from the sections Aigeiros was extracted from M. larici-populina isolates 93ID6 and and Tacamahaca and their hybrids (Frey et al. 2005). During 98AG31 (Table 1) using DNeasy PlantMini Kit (QIAGEN) the last few decades, thispathogen has caused important after grinding5 mg of urediniospores with glass beads economic losses to poplar cultivation in Europe - mainly (peiet al. 1997). A first library was constructed accordirtg based onPopulus x euramericana and Populus x interamericana to Edwards etai. (l996) using Hybond membranes with hyhrids- due to the breakdown ofseveral major resistance two hybridization rounds. A second enriched libraryusing genes (Pinon & Frey2005). RandoII\ amplified polymorphie biotinylated oligoprobes and sh'eptavidin-coated magnetie DNA markerswere developed forAI larici-populina to com­ beads was constructed following the protocol of Dutech paregenetiediversitybetweenstrains occurringoncultivated et al. (2000) to increase the number of loci. Both libraries hybrid poplarstands and wild riparianPopulus nigra stands were enrichedwith (TC)IO and (TG}10 motifs and doned (Frey et al. 2005). However, observedheterozygositycannot using a TOPO TA Cloning Kit (Invitrogen). In the second be detected in this dicaryotic fungus with these dominant library, a colony bIot screening according to Estoup et al. markers, which prompted a searchfor codominant micro­ (1993) and a polymerase chain reaction (PCR) size screen­ satellite markers. Microsatellite markers have alreadybeen ing llSing vector's primers were performed to select clones developed on several Basidiomycetes including !TIst fungi larger than 500 bp and containing microsatellite motifs. In (e.g. Enjalbert et al. 2002). the first library, 104 clones were selected and sequenced, 5ince mierosatellites seem to be less frequent in fungal and 20 contained a microsatellite motiflarger than 20 bp in genomes than in other organisms (Lim et al. 2004), the length. In the second library, 3930 clones were screened strategyofisolatingloci in M. larici-populina was to bulld an and 978 gave a positive response for microsatellite presence. After sizescreening, 96 clones were sequenced and 17 were Correspondence: P. Frey, Fax: +33383394069; chosen for primer design. Primers were designed using E-mail: [email protected] PRIMER 3 software (Rozen & Skaletsky 2000) and tested

© 2006 Blackwell Publishing Ltd

30 Développement de marqueurs microsatellites

PRIMER NOTE 61

Table 1 lsolates ofMelampsom spp. used in the studyc Pathotypes of Mela1"lfpsoyalarici-populina determined as describedin Pinon & Frey (2005)

Species Isolate code Pathotype Original hast Collection site Year

Melampsom larici·populina World subpanel 97Al 0 Populus x eummericana Morocco 1997 97A3 3-4 P.nigra New 2ealand 1997 97110 3-4 P. x euramericana 1-488' South Africa 1997 97EA2 4 Populus sp. China 1997 97C3 1-3-4-5-6-7 P. X interamericana 'Beaupré! UK 1997 98AE3 2 Populus sp. Finland 1998 9901 3-4 P. trichocarpa 1celand 1999 00A19 0 P. X euramericana '1-488' Chile 2000 98ARI 1-3-4-5·7-8 P. x interamericana 'B-71085/Al' Belgium 1998 95USI 3-4 P. nigra 1talica' USA 1995 Regional subpanel 94ZZ5 1-3-4-5-7 P. xeuramericana 'Ghoy' N France 1994 95AAH12 1-3-4-5~6-7 P. xinteramericana 'Beaupré' NE France 1995 95À'P5 1-3-4-5·6-7 P. X interamericana 'Beaupré' SWFrance 1995 95XW2 1-3-4-5·7 P. X intemmericana 'Boelare' WFrance 1995 98AG31 3-4-7 P. x interamericana 'Beaupré' N France 1998 98A02 3-4-7 P. x·itzteramericana "Beaupré' SWFrance 1998 OOMlO 3-4-7-8 P. X ïnteramericana '69038-1' NE France 2000 011.31 2-3-7 P. X euramericana 'A4A' C France 2001 01L66 1-2-3-4-5-7 P. xeuramericana 'Laxo 7' NE France 2001 02A033 4 P. nigra SE France 2002 Local subpanel 99I006 3-4 P. xeuramericana 'Robusta' NE France 1999 99I009 2-4 P. x euramericana 'Robusta' NE France 1999 99I036 3-4-7 P. x euramericana 'Rabusta' NE France 1999 991046 2-3-4 P. X euramericana 'Rabusta' NE France 1999 99I056 1-4-5 P. x euramericana 'Robusta' NE France 1999 99I061 1-3-4-s..6 P _X euramericŒna 'Robusta' NE France 1999 99I094 1-2-34-5-7 P. xeuramericana 'Robusta' NE France 1999 99I108 4 P. xeuramericana 'Robusta' NE France 1999 99I179 3-4-5 P. X euramericana 'Robusta' NE France 1999 99I192 4-7 P. xeuramerical1a 'Robusta' NE France 1999 93ID6 3-4 P. x eU1'americana 145-51' NE France 1993 Melampsom aIlii-populina 9821 Arum sp. SE France 1998 98Z3 Alliumsp. SE France 1998 98Z4 Alliumsp. SE l'rance 1998 98Z5 AlHumsp. SE France 1998 98Z6 Arum sp. SE France 1998 98Z9 lvIuscari comosum SE France 1998 98210 Alliumporrum SE France 1998 89Al P. xeuramericana 'Altichiero' SWFrance 1989 96B6-1 Allium vineale NE France 1996 96M24-2 Muscari comosum NE France 1996 97M2 Allium sphaerocephalum NE France 1997 97Nl Arum macu/atttln NE France 1997 94HY8 P. nigra SE France 1994 94N7 P. nigra SE France 1994 91E4 P. x euramericana 'Robusta' WFrance 1991 91L5 P. x interamericana 'Beaupré' WFrance 1991 Melampsora medusae 88MM2 P. x interamericana 'Beaupré' SW France 1988 f. sp. deltoidae 97CNI P. x interamericana 'Boelare' SWFrance 1997 98Bl P. X interamericana '87002-21' South Africa 1998 98Dl P. xeuramericana '5006' South Africa 1998 99A3 P. X interamericana 'Hoogvorst' SWFrance 1999 9912 P. X interamericana 'Hazendans' NE France 1999 0121 P. deltoides Canada 2001

©2006 Blackwell Publishing Ltd, Molecular Ecology Notes, 6, 60-64

31 Chapitre 1

62 PRIMER NOTE

Table 1 Ccmtinued

5pecies lsolate code Pathotype Original host Collection site Year

01Z2 P. deltoides Canada 2001 01Z4 P. deltoides Canada 2001 0125 P. de ltoides Canada 2001 01211 P. X interamericana 'U nal' Canada 2001 02A25 P.deltoides Canada 2002 02AZ6 P. deltoides Canada 2002 02AZ7 P. deltoides Canada 2002 02AZ8 P.deltoides Canada 2002 02AZll P. X jackii Canada 2002 Melampsora laricMremulae 01Fl P.tremula NE France 2001 Melampsora rostrupii 01Gl Mercurialis perermis NE Franee 2001 Melampsora pinitorqua 0051 P. tremula 5W France 2000 0052 P.tremula 5 France 2000 0053 P.II·emula C France 2000

on a panel of30 M. larici-populina isolates comprisingthree Loci pMLP12 and ftMLP13 were isolated from the same nested geographieal (world, regional, local) subpanelswith clone sequence and arethusphysicallylinked, with a 90-bp high pathotype diversity (Table 1). interval betweenthe twomicrosatellite motifs. Nevertheless, PCRwasperforrned usinga PTC-200 Peltier ThermalCycler no significant linkage disequilibrium CP < 0.01) was detected (MI Research) with 5 min at95 oC followed by 40 cycles among ail pairs of loci for the 30 isolates of the test panel, of 60 s at 94 oC, 90 s at annealing temperature and 60 s at according to an exact test performed with GENEPOP 3.4. 72 oC and a final extension step of 30 min at 60 oc. PCR Ail loci were tested for cross-spedes amplification against was carried out in 20 !J.L final reaction volumes containing M. allii-papulina (16 isolates), M. medusae f. sp. deltoidae (16 15 ng template DNA, 2 pL of lOx reaction buffer, 1.5 mM isolates), M. larici-ttemulae (l isolate), .M. rostrupii (l isolate) MgOz' 0.2 mM dNTP, 0.5 U Taq polymerase (Sigma) and and M. pinitorqUil (3isolates). Six loci resulted in global or 0.2!J.M forward and reverse primers. Out of 37 loci tested, partial amplification with some related species, but very 22failed to have an interpretable pattern of migration few intraspecific polymorphisms were detected (Table 2). ona2% agarose gelbecause of multiple bands orunexpected allele size. Forward primers ofthe remaining 15 loci (Table 2) Acknowledgements were labelled to allow size and dye multiplexing. peRpro­ ducts were separated, sized and analysed on a CEQ 8000 We thank Christine Delaruelle, Anllegret Kohler, Céline Lalanne, Genetic Analysis System (Beckman Coulter). Patrick Léger and Mart; na Peter for their advice and help. This work Out of 15 loci tcstcd, 13 were found to be polymorphie was partly funded by grants from GlP ECOFOR and from INRA. with three to eight alleles(Table 2). Expected and observed heterozygosities as weB as a test for Hardy-Weinberg References expectation were computed using CENEPOl' 3.4 (Raymond & Rousset 1995). Departure from expected results was Altschul 5F, Madden TL, 5ehiiffer AA et al. (1997) Gapped BLAST signifieant for eight loci, whichcould he explained by either and PSI-BLAST: a new geller~tion of protein database se~reh programs. Nucleic Acids Research, 25, 3389-3402. a nonequilibrium state of the population, the sampling Dutech C, Amsellem L, Billotte N, Jarne P(2000) Characterization structure or a low frequency ofnull alleles. Locus J.lMLP22 of (GA). microsatellite loci using an emiehment protoeol in the wassuspected to be uniparentally inherited hecau'se ofthe neotropical tree spedes Vouacapoua americana. Molecular Ecology, totallack of heterozygotes within the panel. The sequence 9,1433-1435. adjacent to this locuswas shown ta exhibit 56% identity Edwards KJ, Barker JRA, Daly Ar Jones C, Karp A (1996) Micros­ (29/51amino acids) with the5' end of the cytochrome atellite libraries enriched for several microsatellite sequences in c oxydase subunit 2 gene of the mitochondrion of the plants. BioTec/miques, 20, 758-760. Enjalbert J, Duan X, Giraud T et al. (2002) Isolation of twelve Basidiomycete Crinipellis perniciosa using NCBI's BLA8TX mierosatellitelod, using an enriehment protoeol, in the phyto­ program (Altschul et al. 1997), strongly suggesting that pathogenie fullguS Puccinia striiformis f. sp. tritici. Molecular Ecology !J.MLP22 was a rnitochondriallocus. Allother loci did not Notes, 2, 563-565. show anysignifieanthomology with nudeotide or protein Estoup A, Solignae~ Harry M, CornuetJM (1993) Characterization databases. of (GT)" and (CT)" mierosatellites in two insect species: Apis

© 2006 Blackwell Publishing Ltd, Molecular Ecology Notes, 6, 60-64

32 ® N <:> <:> Cl'> ~ ~ ~. '"d ~ Table 2 Characteristics of micrqsatellite loci isolated from Melampsora larici-papulina, Ta' armealing temperature; HF! unbiased expected heterozygasity; HO' abserved heterazygasity; far ~ each cambination of related species and locus, ratia between positive isolates and total nurnber of isolates tested, as weil as apparent fragment size when cross-amplification occurred, ~ were indicated, *$ignificantdifference between HE andHo (P < 0,05) t""""'

fi. Milampsora ~ GenBank Expected T. Nfelampsora mmu,,", Melm"p"". Mela:mpsara lvlelampsora [ Locus Accession nos Primer sequences (.5'-3'j Repeat motif of cloned allele ,ize (bp) ('C) Number of alle)e, (size in bp) H. Ho allii~papu1il1a f. sp. delto1dae lm'lC'{~tremuÙle rastrupu' pinitarqua

!f DQ059602 FI 'fCAN:llACCAAACfèAATCMAC (CAJ,C(CA),A(CA), 171 60 4 0.40 0.47 0/16 0/16 0/1 0/1 0/3 t!i !J,MLP09 8 RI CA.'IGCCA.AACCATAC'I'2C'IG (159,165,167,169) !1MU'12 DQ059603 F'l. 'roC=.oI\AACAA()J>.OO1'.Ai'G (MGI,( .. J(AAG),( ...) 24'0 60 7 0.70 0.Q3 0/16 0/16 0/1 0/1 0/3 ~ RI ACC'!T1'G'fCGACC'J\ACAACfè (AAcIi,,(...)(ATG),(."J\G), (239,242,245,247, 24S, 250, 266) ~ !J,MLP13 DQ059604 FI G==C'AAIoOG'ro ('ro)" 187 60 4 0.45 0.47 0/16 0/16 0/1 0/1 0/3 ~ RI Gel''!'AAAGCq;:croA'!'I'ITC /189,191,193,203) CT'ITCPJl.GÇTCM.'I'('~ 181 60 3 0.07 0.07 0/16 0/16 0/1 0/1 0/3 P' llMJ-P20 DQ059605 Pl ('!.G)10 Cl'> RI JW>,CCAAGCAAGCATAœAC (177,179,181) <:> 1 llMJ-P22 DQ059606 FI OOCA'I'K,'TATITATrr'AAA.CACl\.œ (TA,lg( ..)(TAATAG)n( ..)('ro)., 315 60 4 0.62 0.00' 16/16 16/16 0/1 1/1 0;3 ..Cl'> RICl:JA~=TAC=Cfè (314318,324,330) 280 bp 291 bp 274 bp !J,MLP24 DQ059607 FIGA'IC'ACGCCCA'IG:::'lTI'AAC ('ro),( ..)(N), 154 60 1 0.00 0.00 0/16 0/16 0/1 0/1 0/3 R1ATI'~1GAŒ (156) !J,MLP27 DQ059608 F1 CA'R3C'ThA'l'IDTA'1''1'GAGCTGTATffi (TC),TI"I(TC), 250 60 4 0.49 0.13" 13/16 0/16 0;1 1/1 0/3 Rl·'I\X}A'1'GACDA'TGA'IG1~ (240,242,250,252) 228-230 bp !J,MLP28 DQ059609 Pl. ATCCCA'IG:GAA'IC'CG..OJi.A.'IG (CP.cCA),(...)(CAœ.A), 459 60 6 0.29 0.21" 0/16 0/16 0/1 0/1 3/.3 317-,31 RIGCTGA===J:;TccAI.>.AA'I'CA'fCAC (CP.CCA), 298 60 ,3 0.07 0.04' 0/16 0/16 0/1 0/1 RI GCTOX'J'I.GT'IG'ImTAG'mAG (295,300,310) 249 63 5 0.25 0.13' 0/16 0/16 0/1 0/1 0/3 tJ !J,MLP30 DQ059611 FI'IGA'It»:n'TAC.A.'IGA'illP..TI'CC (=1UCO,(...) ~. RI CAACACACAACAACACACAAT'fC ('ID'ro'IDCO( ...)(']1}']1}1GCO (247,249,256,263,314) 0/16 0/1 0/1 0/3 l'Ml.P31 DQ059612 FI CCAŒACAGJ\GAOOATATAGT'IG (G1l;,GC'IG(GCO, 194 .54 6 0.22 0.20 0/16 cs-"" RI CTrTCCCACACTroTTI'lœ (1~1, 183, 184, 187, 190, 192.1 0.82 1/16 1/1 0/1 2/3 !J,MLP32 DQ059613 F2==TTAAGA== ('re)" 132 60 8 0.93' 0/16 118bp RI AAAGCCAACMllGATGAÇC1'G (116,122,124,126,128,136,138,140) 136 bp 118 bp ~ 0/16 0/1 0/1 0/3 !J,MLP34 DQ059614 F3 GAGAAA'fCGAAŒCCPGMG (AGJ"M(h:i), 298 51 1 0.00 0.00 0/16 ~ R2'ICA1>.Al!GllATmmA= (268) "" 0.59 0.57' 0/16 0/16 0/1 1/1 0/3 !J,MLP36 DQ059615 FI TI''rr'.AAAAAGTAA~'ro'1G (AG)" 240 50 6 """ 228-236 bp ~ RI 'l'C'AmTI'AGCTITTCG'I'l'\3.J (226,228, 230, 232, 236, 238) 0.50 0.63' 0/16 0/16 1/1 0/1 2/3 ~ 1'MLP37 DQ059616 FI CCAQ:1GTI'(,."IGAJl.GT1."G'I'A'l"K'· ('IC.}10 224 60 3 '1l 223-229 bp 225-231 bp RI ACT~CGTK'GA.'K'ACC (222, 224, 226) ;;::J ~ Z s:: :::: tr1 ;;::J ""~ Z ~ 0..., ~. tTl ~ 0'1 ~ '" :::::: w &" w Chapitre 1

64 PRIMER NOTE

mellifera and Bombus terrestrls. Nue/de Acids Research, 21, 1427­ Pinon J, Frey P (2005) Interactions bettveên poplar clOnêS and 1431. Metampsora populations and their implications for breeding Frey P, Gérard P, Feau N, Husson C, Pinon J (2005) Variability for durable resistance. In: Rust Diseases of Wittow and Poplar (eds and population biology of Melampsora rusts on poplars. ln: Pei MH,McCrackenAR), pp. 139-154.CAB International, Wall­ Rust Diseases of WiIlo!O and Poplar (eds Pei MH, McCracken AR), ingford, UK. pp. 63-72. CAB International, Wallingford, UK. Raymond M, Rousset F (1995) GENEPOP (version 1.2): population Lim S, Notley-McRobb L, Lim M, Carter DA (2004) A comparison genêtics softwarê for exact tests and ecumenicism. Journal of of the nature and abundance of microsatellites in 14 fungal Heredity, 86, 248-249. genomes. Fungal Genetles and Blology, 41,1025-1036. Rozen S, Skaletsky H (2000) PRIMER 3 on the W'NW for general Pei MH, Whelan M}, Halford NG, Royle DJ (1997) Distinction users and for biologist programmers. In: Bioinformatics Methods betweenstem-and leaf-infecting forms of Melampsora rustonSaUx and Protocols: Methods in Molecular Bi%gy (eds Krawetz S, vlminalis using RAPD markers. Mycological Research, 101, 7-10. Misener 5), pp. 365-386. Humana Press, Totowa, New Jersey.

© 2006 Blackwell Publishing Ltd, Mo/ecular Ecology Notes, 6, 60-64

34 Développement de marqueurs microsatellites

III) Conclusion

Grâce à la constitution de plusieurs banques génomiques enrichies en différents motifs microsatellites, nous sommes parvenus à développer 15 marqueurs microsatellites chez M larici-populina. Le polymorphisme de ces marqueurs a été testé sur un panel d'isolats représentatif de populations locale, régionale et mondiale. Treize de ces marqueurs se sont révélés polymorphes. La transférabilité des loci microsatellites a également été testée sur cinq espèces proches: M allii-populina, M medusae f. sp. deltoidae, M larici-tremulae, M rostrupii, et M pinitorqua. Peu de loci ont permis l'amplification de l'ADN de ces espèces génétiquement proches de M larici-populina. De plus, ces quelques loci transférables se sont révélés peu ou pas polymorphes. Parmi les marqueurs microsatellites isolés, un locus est supposé être situé sur l'ADN mitochondrial. En effet, la séquence d'acides aminés traduite à partir des séquences flanquantes du fragment d'ADN séquencé montre des similarités importantes avec celle d'une enzyme de la chaîne respiratoire (cytochrome oxydase 2) des champignons. L'obtention de marqueurs microsatellites polymorphes permet d'envisager des études de la diversité génétique de M larici-populina. Une première approche qui nous a semblé essentielle fut de caractériser de manière fine l'organisation de la diversité génétique à différents niveaux hiérarchiques.

35 Chapitre 1

36 Chapitre 2

Etude de la diversité génétique à une échelle spatiale fine à l'aide d'un échantillonnage hiérarchisé emboîté Diversité hiérarchique emboîtée

La coexistence de la reproduction sexuée et asexuée chez de nombreux champignons phytopathogènes peut grandement influer sur la diversité génétique de ces organismes (Brygoo et al., 1998). Lors de la phase épidémique, c'est la reproduction asexuée qui permet la multiplication des individus adaptés et qui engendre parfois une structure clonale des populations de l'agent pathogène. A l'inverse, le caractère obligatoire d'une reproduction sexuée annuelle chez certains champignons peut aboutir à une diversité génotypique élevée. Melampsora larici-populina, agent de la rouille du peuplier, est un champignon hétéroïque et macrocyclique. Celui-ci alterne chaque année entre une phase de reproduction sexuée sur le mélèze et de nombreux cycles de multiplication asexuée sur le peuplier. Lors de la phase de reproduction asexuée, d'énormes quantités d'urédospores sont produites et dispersées par le vent. Pour tenter d'appréhender l'importance de chacun des modes de reproduction et de cerner l'organisation fine de la diversité de M larici-populina, nous avons réalisé un échantillonnage hiérarchisé et emboîté à quatre niveaux (feuille, rameau, arbre, site). Trois sites ont été choisis en France: deux sont des pépinières expérimentales (Amance, Meurthe­ et-Moselle et Charrey-sur-Saône, Côte-d'Or) constituées de nombreux clones de peupliers dont certains possèdent des résistances qualitatives à M larici-populina. Le troisième site se situe dans la vallée de la Durance (Prelles, Hautes-Alpes) et est constitué d'une ripisylve sauvage de peupliers noirs. Sur chaque site, cinq arbres ont été sélectionnés, sur chaque arbre trois rameaux ont été échantillonnés, sur lesquels trois feuilles ont été prélevées et sur chaque feuille trois urédies ont été isolées. Un aspect souvent négligé dans les études de la diversité des champignons phytopathogènes est l'évolution de la diversité génétique au cours de la phase épidémique. Deux forces s'opposent alors: d'une part la multiplication asexuée des individus les plus adaptés qui peut entraîner une augmentation de la clonalité dans la population étudiée, et d'autre part, la migration d'individus venant de populations proches qui est une source de nouveaux génotypes. Le second objectif de cette étude est d'évaluer l'évolution de la diversité génétique des populations de M larici-populina au cours de la phase épidémique. Pour cela, sur les sites de Prelles et de Charrey, une population a été échantillonnée en début d'épidémie et une autre en fin d'épidémie. Toutes les populations ainsi échantillonnées ont été analysées à l'aide des marqueurs microsatellites décrits dans le chapitre 1.

Cette étude est présentée sous la forme d'une publication en préparation: Barrès B, Dutech C, Andrieux A, Pinon J, Frey P, Nested hierarchical analysis of population genetic structure ofthe poplar rust fungus, Melampsora larici-populina.

37 Chapitre 2

Nested hierarchical analysis of population genetic structure of the poplar rust fungus, Melampsora larici-populina.

Benoît BARRÈS,* Cyril DUTECH,t Axelle ANDRIEUX,* Jean PINON* and Pascal FREY*

*INRA Nancy, Equipe de Pathologie Forestière, 1136 Interactions Arbres-Microorganismes, IFR 110, 54280 Champenoux, France tINRA Bordeaux, Equipe de Pathologie Forestière, UMR BIOGECO, Domaine de la Grande Ferrade, BP 81,33883 Villenave d'Ornon Cedex, France

Keywords: asexual reproduction, c1ona1ity, genetic diversity, microsatellites, Populus nigra, spore dissemination

Correspondence: P. Frey. Fax: +33 383 394069; E-mail: [email protected]

Running title: Genetic diversity ofM larici-populina

38 Diversité hiérarchique emboilée

Abstract

A nested hierarchical sampling strategy was used to give a first description of the population genetic structure of Melampsora larici-populina using codominant markers. Four nested hierarchicallevels were taken into account: leaf, twig, tree, and site, and a total of 641 isolates were analyzed. A high level of genetic diversity was found with more than 90% of the variation within 1eaves, which was in agreement with an obligate annual stage of sexual reproduction. M larici-populina populations were also shown to be significantly structured at the site and tree levels. The low differentiation between populations at the site level (FST = 0.053) suggested a high degree of gene flow at the inter-regional scale. At a fine scale, our results showed that clonality is mostly located at the leafand the twig levels, especially for the populations collected from the natural stand. Therefore, to avoid sampling same individuals originating from a short distance dispersal of asexual reproduction, our results suggest not to sample several uredinia on the same twig.

39 Chapitre 2

Introduction

Combination or alternation of sexual and asexual reproduction could have a great impact on genetic diversity (Balloux et al., 2003). For example, genotypic diversity can be strongly reduced during burst of clonaI reproduction (e.g. Maynard Smith et al., 1993). However, diversity depends mainly on the proportion of asexual vs. sexual reproduction, coupled with the dispersal ability of spores. In extreme cases where sexual reproduction is unknown, as for the wheat stripe rust fungus Puccinia striiformis, genotypic diversity was shown to be very low even at the continental scale (Hovm011er et al., 2002). The amount and the importance of sexual vs. asexual reproduction are specific of each species and depend on three main features: biology, space and time (Taylor et al., 1999). For example, many plant pathogenic fungi are subject to seasonal episodes of sexual reproduction which alternate with burst of clonaI reproduction. Biological key factors are variation within the mating system and interactions with the host plants (e.g. distance between the aecial and the telial hosts for the rust fungi (Agrios, 1988)). Concerning spatial aspects, the perception of reproductively isolated units and reproductive mode could be greatly affected depending on the geographical scale studied. Finally, time is also an essential parameter affecting the relative importance of sexual vs. asexual reproduction because of successive clonaI reproduction cycles during the growmg season. In order to understand and quantify the relative importance of each reproductive system in a fungal pathogen, the structure of the genotypic diversity should be determined. Even if there is not a unique way to design a sampling strategy, because it greatly depends on the questions being addressed, a multiple nested hierarchical sampling scheme is a good beginning to detect the main patterns of diversity and the most important genetic structure (McDonald, 1997). Several population genetic studies of airborne plant pathogens have used such a nested hierarchical sampling strategy, as for example for the wheat pathogen Mycosphaerella graminicola (Linde et al., 2002) or the poplar pathogen Mycosphaerella populorum (Feau et al., 2005). A nested hierarchical sampling strategy could be completed by a prior knowledge ofthe population structure ofthe organism studied or ofrelated species. In the case ofthe rust fungi Melampsora spp., population genetic studies have been already conducted on several species. Diversity of Melampsora fini (flax rust) populations was measured with both RFLP and isozyme markers (Burdon & Roberts, 1995). The population structure of the poplar rust fungus Melampsora medusae f. sp. deltoidae was studied using two SSCP markers by Bourrassa et al. (1998). Recently, a very high genotypic diversity was shown in 13

40 Diversité hiérarchique emboîtée

populations of the Eurasian poplar rust fungus Melampsora larici-populina using both virulence and RAPD markers (Gérard et al., 2006). This study provided a very useful first estimation of the organization of the genetic diversity of M larici-populina populations and suggested that a smaller sampling scale could be ofinterest. The genetic structure of plant pathogens should change during the growing season. Epidemiology studies have shown the succession of sexual and asexual reproduction cycles during the year, especially for phytopathogenic fungi. Generally, new populations are founded by sexual spores at the start of the growing season and asexual multiplication occurs during the whole growing season. In the second part of the epidemic cycle, individuals with greatest reproductive success should be amplified and take advantage on other individuals. However, even if the asexual multiplication phase is often supposed to be responsible for the disease severity, little is known about the evolution of both allelic and genotypic diversity during the epidemic. Furthermore, recent findings have shown the importance of primary inoculum and the overestimation of the contribution of asexual reproduction to the epidemic (Gobbin et al., 2005). Temporal changes of genetic diversity between the start and the end of an epidemic cycle could be informative about the effect of asexual reproduction on the organization ofthe population diversity.

According to the classification established by McDonald and Linde (2002), the poplar rust fungus M larici-populina belongs to the category of plant pathogens with the highest evolutionary risk because of (i) its mixed reproduction system which alternates obligate sexual reproduction on larches (Larix spp.) and exponential asexual multiplication on poplars (Populus spp.), (ii) its large effective population size and (iii) its probable ability to mutate. Indeed, M larici-populina is a heteroic and macrocyclic rust fungus, i. e. it has a complex life cycle with five kinds of spores and needs two obligate plant hosts to complete its life cycle (Frey & Pinon, 2004). The fungus overwinters as teliospores on dead poplar leaves. In early spring, caryogamy and meiosis occurs, and haploid basidiospores are released, which infect freshly emerged larch needles. After spermatization of spermogonia on larch needles, dicaryotic aecia are formed and release wind-bome aeciospores, which infect poplar leaves. During the whole growing season, M larici-populina reproduce asexually, each uredinium producing thousands of dicaryotic urediniospores each day (Dowkiw et al., 2003), which can be wind-dispersed on very large distances and infect new poplar leaves. In addition, commercial poplar cultivation exhibits two features which can increase the rate of evolution of M larici-populina by imposing a strong and repetitive directional selection on pathogen

41 Chapitre 2

populations. First, poplars are cultivated in large monoclonal stands, and second, unlike annual crops, the cultivation cycle ofpoplar lasts for a minimum of 15-20 years. For aU these reasons, it appears very difficult. to breed for durable qualitative resistance to M larici­ populina (Dowkiw & Bastien, 2004). Moreover, the rapid and successive breakdown of aU of the qualitative resistance genes selected so far and released in the past 30 years tends to demonstrate the inefficiency ofsuch a strategy (Pinon & Frey, 2005). Severe economic losses are caused by poplar rust, and the increasing importance ofpoplar for wood industry or as a potential source of renewable energy prompted the search of new control strategies of rust epidemics. A first strategy is breeding poplar cultivars with a greater quantitative resistance, which is supposed to be more durable (Dowkiw et al., 2003). A second strategy is the deployment in time and space of the available qualitative resistances with a better knowledge of the dispersal and the evolutionary potential of the pathogen (McDonald & Linde, 2002). This approach requires population genetics studies to estimate parameters such as genetic diversity, genetic structure, migration rates and distances of dispersal. The first issue to address before performing population genetics studies is the kind ofmarkers to be used. Microsatellites are one ofthe most powerful Mendelian markers (Jarne & Lagoda, 1996): they are neutral, commonly polymorphic and codominant. The dicaryotic state of urediniospores of M larici-populina prompted us to develop this type of markers (Barrès et al., 2006). There were three main aims to this study. The first aim was to describe the genetic diversity of the pathogen at different spatial scales, from leaves to regions, using a nested hierarchical sampling scheme. In order to compare the spatial distribution of genetic diversity in two contrasted environrnents M larici-populina populations were collected in two experimental poplar nurseries and in one natural Populus nigra stand. We also wanted to test the significance of each level on the structure of the population and to quantify the relative importance of sexual vs. asexual reproduction. The second objective was ta test the temporal stability ofthe genetic structure across an epidemic cycle. In this goal, two sites were sampled at the beginning and at the end of the epidemic season. The third aim was to test the hypothesis of a limited dispersal of sexual spores produced on larches that would have resulted in a genetic structuration ofpopulations on the adjacent poplars.

Materials and Methods

Sampling strategy

42 Diversité hiérarchique emboîtée

A nested hierarchical sampling scheme was used to explore the genetic structure and diversity of M larici-popu1ina populations. Populations were collected in three locations (Figure 1, Table 1): two experimental poplar nurseries located at Amance (AMN, INRA poplar nursery, Champenoux, Northeastern France) and Charrey (CHR, Afocel poplar nursery, Charrey-sur­ Saône, Eastern France) and a natural riparian stand of black poplar (P. nigra) located at Prelles (PRE, Durance River valley, Alps, Southeastern France). Both experimental nurseries contained many different poplar clones, sorne with qualitative resistances. Furthermore, larches (Larix decidua), had been planted within the poplar rows to promote sexual reproduction and early infection, with intent to optimize the pathogen's diversity for rust resistance tests (Pinon & Frey 2005). In both nurseries, M larici-populina populations were collected on P. x euramericana 'Robusta', a cultivar with no qualitative resistance known. In the riparian stand, populations were collected on a random selection of P. nigra trees, no information being available on their genetic diversity. No qualitative resistance to M larici­ populina has been detected so far in P. nigra (Pinon & Frey, 2005). This riparian site was also chosen because poplars and larches (L. decidua) were naturally mixed. Sampling was conducted on two dates at Charrey (CHR-A and CHR-B) and Prelles (PRE-A and PRE-B) in the aim ofmonitoring the evolution of gene and genotype variation over the annual epidemic. At each location, five trees were selected and marked within an approximately 400 m2 area. Distances between trees were measured trunk to trunk with a decameter. Three twigs on each tree were randomly selected and then three leaves on each twig were collected. The collected leaves were stored in an ice box until arrivaI to the laboratory. Three uredinia were randomly selected from each leaf. Approximately 4 mm2 leaf pieces bearing one individual uredinium were cut out with a sterile scalpel and stored at -20°C in Eppendorf tubes until DNA extraction.

DNA analysis

Recently, microsatellite markers were developed for or transferred to M larici-populina (Steimel et al., 2005; Barrès et al., 2006). Since microsatellites are species-specific, they allow direct DNA extraction from single uredinia cut out from infected poplar leaves, as described by Bourassa et al. (2005). DNA was extracted from single uredinia using DNeasy® 96 Plant Kit (Qiagen). We followed the Fresh Leaves protocol (DNeasy® 96 Plant Handbook,

43 Chapitre 2

September 2002) except that samples were disrupted with one tungsten carbide bead during 2 X 1 min instead of2 X 1.5 min at 30 Hz. DNA was eluted in a final volume of200 ~I. Eleven microsatellite loci were chosen among those developed by Barrès et al. (2006). PCR were performed individually in a PTC-200 Peltier thermal cycler (Ml Research) using conditions previously described (Barrès et al., 2006), except for locus ~MLP31 where the PCR mix was modified as follows: 15 ng template DNA, 2 ~l of 10X reaction buffer, 3 mM MgCh, 0.7 ~g/~l BSA (Sigma), 0.2 mM dNTP, 0.5 U Taq polymerase (Sigma) and 0.2 ~M forward and reverse primers in a 20 ~L final reaction volume. To allow size and dye multiplexing, forward primers were labeled with three different dyes (Proligo), D2 for ~MLP13, ~MLP22, ~MLP27 and ~MLP37, D3 for ~MLP20, ~MLP28 and ~MLP30, and D4 for ~MLP09, ~MLP12, ~MLP31 and ~MLP36. PCR products were separated, sized, and analyzed on a CEQTM 8000 Genetic Analysis System (Beckman Coulter). In order to reduce the number of analyses, PCR products were pooled in two sets of loci. Set A was made up of ~MLP09, ~MLP13, ~MLP27, ~MLP30 and ~MLP3610ci with volumes of2, 3, 4, 4 and 4 ~l, respectively, in a 34 ~l final volume whereas ~MLP12, ~MLP20, ~MLP22, ~MLP28, J.lMLP31 and ~MLP37 loci were put together within Set B, with volumes of3,4, 7,4,2 and 8 J.lI, respectively, in a 61 ~l final volume. InternaI size standards of 400 and 600 pb (Beckman Coulter), labeled with Dl dye, were used to genotype Set A and Set B, respectively, in a mixture containing 30 J.lI of Sample Loading Solution (SLS, Beckman Coulter), 0.5 ~l of internaI size standard and 1 J.lI ofeach ofthe marker sets. When chromatograms were ofpoor quality, or when a locus failed to amplify, PCRs were performed again for the entire set. If the analysis failed again, the individual was considered as a missing data. It should be noted that J.lMLP22 was a mitochondriallocus (Barrès et al., 2006).

Data analyses

Identical genotypes were identified using Gimlet version 1.3.3 (Valière 2002) for pooled populations and for each site. Genotypic diversity (G) was estimated at the tree and site levels, following Stoddart & Taylor (1988): G = 1/1JfxCxIN)Z) , where lx is the number of genotypes observed x times in the sample, and N is the number of isolates. The relative genotypic diversity (GIN) and a t-test for the significance of differences between genotypic diversities (Chen et al., 1994) were computed. The insufficient power ofmolecular markers could lead to identify individuals with the same multilocus genotype which are not clonaI. In order to identify multilocus genotypes that are statistically overrepresented assuming panmixia, and

44 Diversité hiérarchique emboîtée

thus that could be considered as belonging to the same clonaI lineage, the method described by Halkett et al. (2005) was used. The probability ofobserving n times a multilocus genotype in a population was computed using MLGsim software (Stenberg et al., 2003). Then the program, using a Monte Carlo simulation method, determines the significance threshold for the probability values for each population, taking into account sample size and allele frequencies. The significance level was fixed to 0.05 in the present study. A clone-corrected dataset was built keeping only one individual per identical multilocus genotypes that were considered as clone, at each site. Genotypic linkage disequilibrium and deviation from Hardy­ Weinberg equilibrium were computed using GENEPOP 3.4 (Raymond and Rousset, 1995) on the clone-corrected dataset, at the site and tree levels, in order to lower clonaI reproduction effects. Because of the physica1 linkage of loci f!MLP12 and f!MLP13 (Barrès et al., 2006), these two loci were replaced by a chimeric new locus, named f!MLP38, reconstructed using PHASE (Stephens et al., 2001). Number of alleles (Na), allelic richness (Ar) and gene diversity (He) were estimated on clone-corrected data for each site and for each poplar tree using FSTAT 2.9.3.2 (Goudet 1995). Permutation tests were also carried out using FSTAT in arder to test if allelic richness and gene diversity were significantly different between M larici-populina populations collected from nurseries vs. the natural poplar stand. Group 1 contained the five trees of PRE-A population and Group 2 contained the 10 trees of AMN and CHR-A populations. AIl the genotypic differentiations and exact tests were conducted using GENEPOP 3.4 (except

when mentioned). Pairwise F ST were estimated using Weir & Cockerham's e (1984). Nei's minimum genetic distances (1972) were computed with the complete dataset, using either the nine nuclear or the single mitochondrial microsatellite markers. Two unrooted cladograms were built using unweighted pair group method with arithmetic averages (UPGMA) and MEGA version 3.0 software (Kumar et al., 2004). Bootstrap values were calculated from 10000 replications over markers. Nested five-Ievel analyses of molecular variance (AMOVA) were conducted using HIERFSTAT package in R (Goudet 2004), which implement the method developed by Yang (1998), on complete and clone-corrected datasets of AMN, CHR-A and PRE-A populations. Variance was partitioned among sites, among trees within sites, among twigs within trees, among leaves within twigs and among individuals within leaves. The effect of each level was tested with G-statistic likelihood ratio tests (Goudet et al., 1996), independently ofthe effect ofthe lower levels in the hierarchy, performing 1000 permutations with HIERFSTAT. A hierarchical analysis of variance was also performed to analyze the partition ofvariance site by site at the four hierarchicallevels in each population.

45 Chapitre 2

Results

Genetic Diversity

In aU, 675 isolates were analyzed. Because of poor DNA quality of a few samples, only 641 readable chromatograms were obtained, from which 445 different genotypes were identified.

Mean numbers of aUeles between populations were similar. The number of genotypes (Ng) was significantly higher in populations coUected from nurseries (AMN, CHR-A and CHR-B) than in populations coUected from the natural stand (PRE-A and PRE-B, Table 2). At the site level, the relative genotypic diversity (GIN) was high in nurseries (mean 0.828 ± 0.028, Table 2), regardless oflocation or sampling date, whereas significantly lower values (P

46 Diversité hiérarchique emboîtée may explain these departures. For example, when tests were carried out at the tree level, the proportion of significant tests decreased to 4.3% (8 out of 136 possible tests). Because of rejection of Hardy-Weinberg equilibrium for sorne populations, testing the population differentiation was carried out at the genotypic level in order to make no assumptions about the equilibrium state ofpopulations.

Distribution ofclonality

No homoplasy was observed between the alleles ofthe chimeric locus IlMLP38, even ifit was theoretically possible (e.g: allele '435' of locus IlMLP38 could either result from the addition ofallele '242' and allele '193' or from allele '248' and allele '187', of loci IlMLP12 and IlMLP13, respectively). Loci IlMLP12 and IlMLP13 were hence successfully replaced by IlMLP38 without affecting clonality rates in populations. A few identical genotypes were detected between sites (Table 3): none between Amance and Prelles, three between Charrey and Prelles and eight between Amance and Charrey. The number and mean size (number of individuals) ofgenotypes sampled several times in a site are much higher in the natural poplar stand than in nurseries (Table 4), in accordance with the significantly (P

Genetic differentiation

Two cladograms were obtained using Nei's minimum genetic distance (Nei, 1972) and the data from either the nine nuclear or the single mitochondrial microsatellite markers (Figure 2). Gnly results obtained with the complete dataset were shown, since the results with the c1one­ corrected dataset were similar. In both cladograms, populations sampled at the same site at two different times were clustered. Nuclear genetic distances between sites were lower (0.004 to 0.026) than mitochondrial genetic distances (0.004 to 0.460) which c1early distinguished the populations from the nurseries and from the natural stand. Even if the genotypic

47 Chapitre 2

differentiation was low between populations, tests performed on the complete dataset were significant (P

Hierarchical diversity

The AMOVA analysis showed that more than 87% ofthe total variance was attributable to the differences within individuals, when computed on the complete dataset as on the clone­ corrected dataset (Table 6). Approximately Sand 6% ofthe variance was found among site in the complete and clone-corrected datasets, respectively. Components of variance were differently distributed in the remaining hierarchical levels between the complete and the clone-corrected datasets. Variance components were weak and almost equaUy distributed within leaves (1.9%), twigs (0.6%), trees (2.0%) and sites (2.6%) in the clone-corrected dataset, whereas most ofthe remaining variance (11.8%) was found among twigs within trees in the complete dataset. The effect of each hierarchicallevel was tested independently using

HIERFSTAT. There was a highly significant effect (P

48 Diversité hiérarchique emboÎtée

populations from the poplar nurseries and from the natural stand were obvious at the twig level: this level had a very significant effect on PRE-A and PRE-B populations and no significant effect on AMN, CHR-A and CHR-B populations. Again it should be pointed out that this result is in agreement with the importance and the distribution ofclones within sites. F-statistics among individuals within twigs resulted in negative values in non clone-corrected

data (F17=-0.555 and Fn=-0.071 for PRE-A and PRE-B, respectively). These results suggest

that there was an heterozygote excess in both populations at the twig level and are lU agreement with an important amount ofclonaI reproduction (Balloux et al., 2003).

Evolution ofdiversity during an epidemic cycle

M larici-populina populations were sampled at two different dates at Charrey (CHR-A and CHR-B) and Prelles (PRE-A and PRE-B) in order to detect a potential evolution of the genetic diversity during an epidemic cycle. Different temporal evolutions were observed in both sites. At Charrey, on 244 genotypes identified, only 17 were common to CHR-A and CHR-B populations (Table 3), illustrating the very high genotype diversity found at this site. However, other parameters suggest that even if different genotypes were sampled, Charrey's population seems genetically stable across an epidemic season: similar values of relative genotypic diversity (GIN, close to 1), of allelic richness and gene diversity were observed for CHR-A and CHR-B populations (Table 2), very low nuclear and mitochondrial genetic distances between CHR-A and CHR-B (Figure 2) as weIl as low and not significant population differentiation if computed with clone corrected dataset (Table 5). Differentiation between tree subpopulations within populations was low and rarely significant. Low differentiations between subpopulations in CHR-A and CHR-B were found (Smax=0.036 and Smax=0.030 in complete and clone-corrected dataset, respectively). At Prelles only 107 different gen-otypes were identified, of which only three were shared by the two populations (Table 3). Although allelic richness and gene diversity had similar values, a two-fold increase ofthe number ofdifferent genotypes (from 38 to 69) and an important and very significant increase of GIN (from 0.207 to 0.355, P

sen UHP N'ANCY 1 Bibliothèque des Sciences 49 Rue du Jardin Botfu~ique - CS 20148 54601 'VILLERS LES NANCY CEDEX Chapitre 2

(P

Discussion

Importance ofsexual reproduction

The total number of genotypes and the genotypic diversity were very high at all the three sites, and more especially in the two poplar nurseries. In addition, more than 90% of the genetic variation was found within leaves, regardless ofthe dataset used. Three uredinia were sampled on each leaf and more than 82% (183/223) ofthese leaves were found bearing more than one genotype. The gene diversity was moderately high in aH the populations and only a few pairs of markers were found in linkage disequilibrium in a few subpopulations. These results highlight the great importance of sexual reproduction in the life cycle of M larici­ populina. This was not very surprising because of the mixing of poplars and larches at all the sites studied, but it reinforces the hypothesis that most of the primary inoculum originates from aeciospores and that sexual reproduction is an obligatory step of the M larici-populina cycle. The alternative hypothesis of winter survival of asexual spores seems very unlikely at the sites studied, and even if it exists, it should have a limited impact on the genetic diversity ofthe populations.

Importance ofasexual reproduction at the site scale

In this study, identification and localization of clonemates at different hierarchical levels showed that most ofthe clonaI lineages where located on a single tree at a site and frequently on a single twig of a tree, especially in the natural poplar stand of Prelles. Values and variation of hierarchical F-statistics between the complete and the clone-corrected datasets tended to confirm this observation. The differences observed in the amount of clonality between the nurseries and the natural stand may be explained by several sampling bias: (i) the

50 Diversité hiérarchique emboîtée surface ofan individual P. x euramericana 'Robusta' leaf is about 10-fold larger than that ofa P. nigra leaf, which drastically reduced the net trapping efficiency at Prelles; (ii) the amount of primary inoculum was probably higher in poplar nurseries because of the spatial arrangement and the number of larches; (iii) as the result of this, the epidemic developed earlier and faster at Charrey and Amance than at Prelles. Taking into account the ability of each uredinium to produce thousands of urediniospores daily (Dowkiw et al., 2003), it was surprising that large clonaI lineages at Prelles (up to 9 isolates in PRE-A population) had not invaded aIl the available foliage two months after, at the second sampling date. Furthermore, a significant decrease of the mean size of clonaI lineages and an increase of the relative genotypic diversity between the first and the second sampling dates was noticed. The lower infection level observed at Prelles compared to Amance and Charrey could either be due to a higher level of quantitative resistance of P. nigra to M larici-populina compared to P. x euramericana 'Robusta' or to dryer climatic conditions during summer. Indeed, leaf wetness plays an important raIe in the infection of poplars by M larici-populina (Pinon et al., 2006). The focal organization of clonality observed at Prelles indicated a preferential short distance dispersal of asexual spores. This focal organization may be explained, at least partIy, by a rain-splashing dispersal, which is characterized by a steeper dispersal gradient, compared to aerial dispersal (Fitt et al., 1989). Existence of rain-dispersal was already reported for other rust fungi like Puccinia spp. (Geagea et al., 1999; 2000) or Hemileia vastatrix (Kushalappa & Eskes, 1989). Furthermore, a similar pattern of clonality was recently observed with the ascomycete Mycosphaerella populorum when fine sampling scale was used (Feau et al., 2005).

Hierarchical structure oiM. larici-populina populations

One of the main aims of this study was to determine the most relevant scale to study the genetic diversity and structure of populations of M larici-populina. It is worth noting that using the complete or the clone-corrected dataset has an important impact on which levels are considered having a strong and significant effect on the genetic structure. The two mains organization levels detected were site and tree. F-statistics among twigs within trees (Frr) sharply decreased from 0.127 to only 0.022 after clone-correction. Significant effect of the twig level depended of whether aIl individuals were considered and was therefore associated with predominance ofshort distance dispersal ofasexual spores.

51 Chapitre 2

Structuration due to sites was moderately high and very significant (Fsr = 0.053 and Fsr = 0.058 for the complete and the clone-corrected datasets, respectively). Similar values were found at a comparable scale using RAPD markers:

statistics among trees within sites were low (Frs = 0.020 and Frs = 0.028 for the complete and the clone-corrected datasets, respectively). Differentiation between pairs of trees within sites varied from no detectable differentiation to high differentiation. In the two populations collected at Prelles, the amount and the clustering of clones have an important impact on differentiation between tree subpopulations: differentiation was very significant when computed on the complete dataset, and became not significant for approximately half of the subpopulation pairs when computed on the clone-corrected dataset. This result highlights the importance ofthe sampling strategy and the effect ofasexual reproduction.

Evolution ofdiversity during the epidemic cycle

The evolution of diversity over the season was highly different at Prelles and Charrey. CHR­ A and CHR-B populations were genetically very close and were not significantly differentiated after clone-correction, indicating a stable state of the population. At the

52 Diversité hiérarchique emboîtée opposite, PRE-A and PRE-B populations were very significantly differentiated, regardless of the dataset used, and a significant increase of aIl the genetic diversity indices was observed between the two sampling dates. Migration from neighboring populations may play an important role in this evolution: the arrivaI of new genotypes can explain the significant increase of the genotypic diversity, gene diversity and allelic richness. Three arguments can partly explain such differences. First, the rapid development of the rust epidemic at Charrey did not favor installation of new migrants between the first and the second sampling dates. Second, compared to Prelles, Charrey's nursery is located in a more fragmented habitat for M larici-populina and is therefore more isolated from others populations. Third, because of the high intensity of the epidemic, the period between the first and the second sampling date at Charrey was two-fold shorter than at Prelles. This may explain why the genetic structure showed a greater variation during the epidemic at Prelles compared to Charrey.

References

Agrios ON (1988) Plant Pathology, 803 pp. Academie Press, London, UK.

Barrès B, Dutech C, Andrieux A, Caron H, Pinon J, Frey P (2006) Isolation and characterization of 15 microsatellite loci in the poplar rust fungus, Melampsora larici­ popufina, and cross-amplification in related species. Molecular Ecology Notes, 6, 60-64.

Balloux F, Lehmann L, de Meeûs T (2003) The population genetics of clonaI and partially clonaI diploids. Genetics, 164, 1635-1644.

Bourassa M, Bernier L, Milligan BG, Hamelin RC (1998) Sympatry between alternate hosts affects population structure of poplar leaf rust. Finnish Forest Research Institute, Research Paper, 712, 65-69.

Bourassa M, Bernier L, Hamelin RC (2005) Direct genotyping of the poplar leaf rust fungus, Melampsora medusae f. sp. deltoidae, using codominant PCR-SSCP markers. Forest Pathology, 35, 245-261.

Burdon JJ, Roberts JK (1995) The population genetic structure ofthe rust fungus Melampsora fini as revea1ed by pathogenicity, isozyme and RFLP markers. Plant Pathology, 44, 270-278.

53 Chapitre 2

Chen RS, Boeger lM, McDonald BA (1994) Genetic stability in a population of a plant pathogenic fungus over time. Molecular Ecology, 3, 209-218.

Chesser RK, Baker Rl (1996) Effective sizes and dynamics of uniparentally and biparentally inherited genes. Genetics, 144, 1225-1235.

Dowkiw A, Husson C, Frey P, Pinon l, Bastien C (2003) Partial resistance to Melampsora larici-populina leaf rust in hybrid poplars: genetic variability in inoculated excised leaf disk bioassay and relationship with complete resistance. Phytopathology, 93, 421-427.

Dowkiw A, Bastien C (2004) Characterization of two major genetic factors controlling quantitative resistance to Melampsora larici-populina leaf rust in hybrid poplars: strain specificity, field expression, combined effects, and relationship with a defeated qualitative resistance gene. Phytopathology, 94, 1358-1367.

Feau N, Hamelin RC, Vandecasteele C, Stanosz GR, Bernier L (2005) Genetic structure of Mycosphaerella populorum (anamorph Septoria musiva) populations in North-Central and Northeastern North America. Phytopathology, 95, 608-616.

Fitt BDL, McCartney HA, Walklate Pl (1989) The role of rain in dispersal of pathogen inoculum. Annual Review ofPhytopathology, 27, 241-270.

Frey P, Pinon 1 (2004) La rouille du peuplier: un pathosystème modèle. Biofutur, 247, 28-32.

Geagea L, Huber L, Sache 1 (1999) Dry-dispersal and rain-splash of brown (Puccinia recondita f.sp. trifici) and yellow (P. striiformis) rust spores from infected wheat leaves exposed to simulated raindrops. Plant Pathology, 48, 472-482.

Geagea L, Huber L, Sache l, Flura D, McCartney HA, Fitt BDL (2000) Influence ofsimulated rain on dispersal of rust spores from infected wheat seedlings. Agricultural and Forest Meteorology, 101, 53-66.

54 Diversité hiérarchique emboîtée

Gérard P, Husson C, Pinon J, Frey P (2006) Comparison of genetic and virulence diversity of Melampsora larici-populina populations on wild and cultivated poplar and influence of the alternate host. Phytopathology, 96, 1027-1036.

Gobbin D, Jermini M, Loskill B, Pertot l, Raynal M, Gessler C (2005) Importance of secondary inoculum of Plasmopara viticola to epidemics of grapevine downy mildew. Plant Pathology, 54, 522-534.

Goudet J (1995) FSTAT (version 1.2): A computer program to calculate F-statistics. Journal of Heredity, 86, 485-486.

Goudet J, Raymond M, de Meueus T, Rousset F (1996) Testing differentiation in diploid populations. Genetics, 144,1933-1940.

Goudet J (2004) HIERFSTAT, a package for R to compute and test hierarchical F-statistics. Molecular Ecology Notes, 5, 184-186.

Halkett F, Plantegenest M, Prunier-Leterme N, Mieuzet L, Delmotte F, Simon JC (2005) Admixed sexual and facultatively asexual aphid lineages at mating sites. Molecular Ec%gy, 14,325-336.

Hovm011er MS, Justesen AF, Brown JKM (2002) Clonality and long-distance migration of Puccinia striiformis f. sp. tritici in north-west Europe. Plant Pathology, 51, 24-32.

Jarne P, Lagoda PJL (1996) Microsatellites, from molecules to populations and back. Trends in Ecology and Evolution, 11, 424-429.

Kumar S, Tamura K,Nei M (2004) MEGA3: Integrated software for Molecular Evolutionary Genetics Analysis and sequence alignment. Briefings in Bioinformatics, 5, 150-163.

Kushalappa AC, Eskes AB (1989) Advances III coffee rust research. Annual Review of Phytopathology, 27, 503-531.

55 Chapitre 2

Linde CC, Zhan J, McDonald BA (2002) Population structure of Mycosphaerella graminicola: From lesions to continents. Phytopathology, 92, 946-955.

Maynard Smith J, Smith NH, O'Rourke M, Spratt BG (1993) How clonaI are bacteria? Proceedings ofthe National Academy ofSciences ofthe USA, 90,4384-4388.

McDonald BA (1997) The population genetics of fungi: tools and techniques. Phytopathology, 87, 448-453.

McDonald BA, Linde CC (2002) The population genetics of plant pathogens and breeding strategies for durable resistance. Euphytica, 124,163-180.

Nei M (1972) Genetic distance between populations. The American Naturalist, 106,283-292.

Pinon J, Frey P (2005) Interactions between poplar clones and Melampsora populations and their implications for breeding for durable resistance. In: Rust diseases ofWillow and Poplar (eds. Pei MH, McCracken AR), pp. 139-154. CAB International, Wallingford, UK.

Pinon J, Frey P, Husson C (2006) Wettability of poplar leaves influences dew formation and infection by Melampsora larici-populina. Plant Disease, 90, 177-184.

R Development Core Team (2004) R: a language and environment for statistical computing. R foundationfor Statistical Computing, Vienna, Austria. URL http://www.R-project.org.ISBN3­ 900051-00-3.

Raymond M, Rousset F (1995) GENEPOP (version1.2): population genetics software for exact tests and ecumenism. Journal ofHeredity, 86, 248-249.

Steimel L, Chen W, Harrington TC (2005) Development and characterization ofmicrosatellite markers for the poplar rust fungi Melampsora medusae and Melampsora larici-populina. Molecular Ecology Notes, 5, 484-486.

Stenberg P, Lundmark M, Saura A (2003) MLGsim: a program for detecting clones using a simulation approach. Molecular Ecology Notes, 3, 329-331.

56 Diversité hiérarchique emboîtée

Stephens M, Smith NJ, Donnelly P (2001) A new statistical method for haplotype reconstruction from population data. The American Journal ofHuman Genettes, 68, 978-989.

Stoddart JA, Taylor JF (1988) Genotypic diversity: estimation and prediction in samples. Genettes, 118, 705-711.

Taylor JW, Jacobson DJ, Fisher MC (1999) The evolution of asexual fungi: reproduction, speciation and classification. Annual Revtew ofPhytopathology, 37,197-246.

Valière N (2002) GIMLET: a computer program for analysing genetic individual identification data. Molecular Ecology Notes, 2, 377-379.

Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358-1370.

Wilkinson AG, Spiers AG (1976) Introduction ofthe popiar rusts Melampsora lartct-populina and M medusae to New Zealand and their subsequent distribution. New Zealand Journal of Science, 19, 195-198.

Yang RC (1998) Estimating hierarchical F-statistics. Evolution, 52, 950-956.

Acknowledgements

We are grateful to Béranger Bertin and Christine Géhin for their help during the collection of M larici-populina populations. This work was supported by INRA and a fellowship of the Région Lorraine.

57 Chapitre 2

Figure Legends

Figure 1: Location ofthe three collection sites in France and detailed maps ofeach site. Black and open triangles correspond to Populus x euramericana 'Robusta' and to P. nigra, respectively.

Figure 2: Unrooted UPGMA c1adograms computed on Nei's minimum genetic distance (Nei, 1972) with the complete dataset using nine nuclear (a) and one mitochondrial (b) microsatellite markers. Bootstrap values computed on 10000 replicates are indicated.

Tables and Figures

58 Diversité hiérarchiQue emboîtée

IAmance AAMN.3

France À ~ AMN-4

AAMN.5 1 AMN'1A A~-2 ~- Amance III L_

Ch-arrey ------~CHR=--l A , CHR·3 1

Charrey III

CHR·2A ÀCHR-4

~ \ A CHR-1 .ml..1

Prelles III Ivll [j.PRE-4

PRE·3 i~ [j. PRE-2[j. [j.PRE-6 [j. 100 km PRE·1 Û PRE·5 '5i1ï

Figure 1: Barrès et al.

59 Chapitre 2

(a)

------_-:..l0:..;:O~I CHR-A ,.,.- 96_1 CHR-B 971 4------::::

0.010 0.008 0.006 0.004 0.002 0.000

(b)

CHR-A l00~ CHR-B AMN PRE-A 4------;c-o PRE-B

0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

Figure 2: Barrès et al.

60 Diversité hiérarchique emboîtée

Table 1: Characteristics ofthe collection sites

Latitude Longitude Elevation Site Host Population ID Sampling date (ddOmm'ss.s") (ddOmm'ss.s") (m)

Amance P. x euramericana 'Robusta' 48°45'017" N 6°20'19.9" E 258 AMN July 2004

Charrey P. x euramericana 'Robusta' 47°05'IIJ" N 5°09'18.7" E 180 CHR-A July 2004

CHR-B August 2004

Prelles P. nigra 6°34'47.2" E Il60 PRE-A August 2004

PRE-B October 2004

61 Chavitre 2

Table 2: Characteristics ofthe M. larici-populina populations and subpopulations at the tree level.

Tree Genotypie Relative Mean number Allelie Gene Population Number of Number of subpopulation diversity genotypic ofalle/es richness diversity FIs ID iso1ates (N) genotypes (Ng) ID (G) diversity (GIN) (Na ± SD) (Ar ± SD) (He ± SD) AMN AMN1 27 22 17.8 0.659 3.33 ± 1.50 1.97 ± 0.64 0.262 ± 0.194 0.152 AMN2 27 27 27.0 1.000 3.11±0.93 2.23 ± 0.73 0.314 ± 0.197 0.016 AMN3 27 25 23.5 0.871 3.22 ± 1.20 2.19±0.78 0.316 ± 0.215 -0.040 AMN4 27 27 27.0 1.000 3.22 ± 2.11 2.24 ± 1.11 0.317 ± 0.261 0.131 AMN5 26 26 26.0 1.000 3.33 ± lAI 2.11 ± 0.80 0.300 ± 0.235 0.004 Population 134 122 108 0.807 4.44 ± 2.19 3.69 ± 1.51 0.313 ± 0.222 0.084 CHR-A CHR1-A 27 26 25.1 0.931 3.00 ± 1.80 2.08 ± 0.98 0.303 ± 0.270 -0.029 CHR2-A 26 25 24.1 0.929 2.78 ± 1.39 1.98 ± 0.80 0.276 ± 0.224 0.069 CHR3-A 27 24 20.8 0.771 3.33 ± 1.50 2.15 ± 0.99 0.309 ± 0.255 0.055 CHR4-A 27 27 27.0 1.000 3.11 ± 1.27 2.07 ± 0.79 0.293 ± 0.227 0.087 CHR5-A 27 26 25.1 0.931 3.89 ± 1.62 2.18±0.91 0.293 ± 0.233 0.158 Population 134 122 109 0.817 4.67±2.24 3.69 ± 1.57 0.297 ± 0.232 0.076 CHR-B CHR1-B 27 25 23.5 0.871 3.33 ± 1.80 2.10 ± 0.87 0.288 ± 0.219 0.124 CHR2-B 27 25 23.5 0.871 3.22 ± 1.92 2.08 ± 0.96 0.292 ± 0.264 0.112 CHR3-B 27 27 27.0 1.000 3.00 ± 1.32 1.93 ± 0.77 0.260 ± 0.233 0.068 CHR4-B 27 27 27.0 1.000 3.33 ± 1.87 2.17±0.99 0.304 ± 0.246 0.039 CHR5-B 27 26 25.1 0.931 3.56 ± 1.81 2.20 ± 0.89 0.300 ± 0.220 0.110 Population 135 125 ]]6 0.860 4.78 ± 2.95 3.73±1.81 0.290 ± 0.227 0.094 PRE-A PRE1-A 20 7 5.0 0.250 2.33 ± 1.87 2.12 ± 1.50 0.278 ± 0.294 -0.029 PRE2-A 20 7 4.9 0.244 2.78 ± 1.30 2.48 ± 1.19 0.340 ± 0.273 0.113 PRE3-A 22 10 7.6 0.344 2.78 ± 1.20 2.39 ± 1.05 0.371 ± 0.263 -0.090 PRE4-A 27 9 4.1 0.174 2.56 ± 1.24 2.15 ± 0.93 0.314 ± 0.247 -0.202 PRE5-A 27 5 4.1 0.151 2.56 ± 1.33 2.56 ± 1.33 0.347 ± 0.279 0.040 Population 116 38 24 0.207 4.56±2.35 4.56 ±2.35 0.353 ± 0.230 0.009 2.71 ± 1.24 0.402 ± 0.239 0.119 PRE-B PREl-B 27 21 16.2 0.600 4.00 ± 2.29 -0.010 PRE2-B 24 13 9.9 0.414 2.89 ± 1.36 2.32 ± 1.17 0.322 ± 0.245 0.366 ± 0.285 0.145 PRE3-B 23 Il 5.7 0.247 3.00 ± 1.80 2.49 ± 1.36 0.374 ± 0.286 -0.026 PRE4-B 23 Il 6.7 0.291 3.11 ± 1.90 2.52 ± 1.33 0.319±0.219 0.099 PRE6-B 25 17 14.5 0.581 3.00 ± 1.23 2.30 ± 0.95 0.365 ± 0.232 0.091 Population 122 69 43 0.355 4.89± 2.98 4.27 ± 2.37

62 Diversité hiérarchique emboîtée

Table 3: Number of identical genotypes (and number of isolates in these genotypes) between pairs of populations (above diagonal) and geographical distance in kilometers (below diagonal).

AMN CHR-A CHR-B PRE-A PRE-B

5 4 0 0 AMN (10) (11) (0) (0) 17 0 0 CHR-A 210 (40) (0) (0) 0 3 CHR-B 210 0 (0) (6) 3 PRE-A 430 270 270 (20)

PRE-B 430 270 270 0

63 Chapitre 2

Table 4: Fine-scale organisation of clonality. For each population, numbers of repeated multilocus genotypes and mean size of these repeated genotypes are given. Percentages (and ratios) of genotypes restricted to only one tree, and within tree, restricted to only one twig were reported for each population, as weIl as the number of repeated multilocus genotypes considered as clones.

Frequency of Proportion of Frequency of Number of repeated repeated Mean size of repeated repeated genotypes multilocus Population ID repeated genotypes multilocus restricted to a genotypes genotypes restricted to a genotypes single twig considered as single tree within a tree clones AMN 8 2.50 62.5% (5/8) 20% (1/5) 118

CHR-A 10 2.20 50% (5/10) 40% (2/5) 2/10

CHR-B 9 2.11 55.6% (5/9) 60% (3/5) 119

PRE-A 24 4.25 100% (24/24) 100% (24/24) 22/24

PRE-B 25 3.12 100% (25/25) 80% (20/25) 22/25

64 Diversité hiérarchique emboîtée

Table 5: Pairwise subpopulation Pm estimated with Weir and Cockerham's 8, for complete (above diagonal) and clone-corrected (below diagonal) datasets between the five M larici­ a populina populations .

AMN CHR-A CHR-B PRE-A PRE-B AMN 0.067 *** 0.066 *** 0.069 *** 0.059 *** CHR-A 0.070 *** 0.009 * 0.052 *** 0.053 *** CHR-B 0.062 *** 0.007 ns 0.044 *** 0.040 *** PRE-A 0.063 *** 0.059 *** 0.036 *** 0.016 *** PRE-B 0.053 *** 0.044 *** 0.031 *** 0.011 ***

a Unbiased estimate ofthe P-value oflog-likelihood based exact test on genotypic distribution using a Markov chain method (dememorization=5000, batches=50, iterations=2000) in GENEPOP version 3.4. Significance levels are indicated by stars (*, ** and *** for P

65 Chapitre 2

Table 6: Analysis of molecular variance (AMOVA) of M larici-populina populations sampled according to a nested five-level hierarchical scheme. The total population is composed of one population of each site (AMN, CHR-A and PRE-A). Variance was partitioned for each hierarchicallevel, corresponding F-statistics were computed, and effect of each level (site, tree, twig and leaf) was tested independently of the effect of the lower levels a in the hierarchy using HIERFSTAT .

Complete dataset Clone-corrected dataset

Hierarchicallevel Variation (%) F-statistics Variation (%) F-statistics

Within individuals 87,5 FIT = 0.125 87,1 FIT = 0.129

Among individuals within -8,3 FIL = -0.105 1.9 FIL = 0.021 leaves

Among leaves within twigs 1,9 FLT = 0.024 ns 0.6 FLT = 0.007 ns

Among twigs within trees Il,8 FTT = 0.127 *** 2,0 Frr = 0.022 ns

Among trees within sites 1,9 FTs = 0.020 * 2.6 FTs = 0.028 **

Among sites 5,3 FST = 0.053 *** 5.8 F~T = 0.058 ***

a Significance levels are indicated by stars (*, ** and *** for P<0.05, P

66 Diversité hiérarchique emboÎtée

Un échantillonnage hiérarchisé et emboîté de populations de M larici-populina nous a permis de décrire la structure génétique de ce champignon pathogène d'une échelle spatiale fine à une échelle inter-régionale. Le premier constat est l'exceptionnelle diversité génotypique rencontrée dans toutes les populations. Au total, 641 individus ont pu être typés et 445 génotypes différents ont été observés. D'après l'analyse de variance moléculaire, plus de 90% de la diversité observée est expliquée au niveau de la feuille. De plus la quasi-absence de déséquilibres de liaison entre les marqueurs microsatellites, ainsi que la très importante diversité génotypique, confirment que la reproduction sexuée est très certainement obligatoire, et que la survie asexuée est improbable sous nos latitudes. En effectuant une analyse hiérarchique de la diversité génétique, nous avons montré que les populations étaient significativement structurées au niveau site et au niveau arbre. La différentiation modérée

entre les sites (P~T = 0.053) indique l'occurrence de flux de gènes importants à l'échelle inter­ régionale. La plupart des génotypes multilocus identiques ont été considérés comme étant issus de la multiplication asexuée. La quasi-totalité des lignées clonales avait une répartition limitée à un seul rameau au sein des populations étudiées. La faible occurrence de lignées clonales d'effectifs importants dans les populations prélevées dans les pépinières, comparées à celles prélevées sur le site naturel, peut sans doute s'expliquer par le moindre avancement de l'épidémie dans le compartiment sauvage, ainsi que par la taille moindre des feuilles de peuplier noir comparée à celle des feuilles du peuplier hybride euraméricain 'Robusta'. Si la diversité génotypique était plus faible dans le compartiment sauvage, nous avons pu constater que la richesse allèlique était elle supérieure. Ce dernier résultat est cohérent avec l'hypothèse d'un centre d'origine de M larici-populina qui se situerait dans une zone de sympatrie naturelle entre P. nigra et Larix decidua, en Europe, par exemple dans l'arc alpin. Le suivi de l'évolution de la diversité entre le début et la fin de l'épidémie semble montrer que la population prélevée dans la pépinière de Charrey-sur-Saône est beaucoup plus stable génétiquement que la population prélevée sur P. nigra à Prelles, qui voit sa diversité génotypique augmenter, vraisemblablement sous l'effet de l'immigration de nouveaux génotypes originaires de populations voisines. Ces résultats montrent que l'échantillonnage d'une population peut être réalisé à l'échelle d'un arbre, en prenant la précaution de prélever les échantillons sur des rameaux différents pour éviter un effet de la multiplication asexuée locale sur la structure de la population. Ils montrent également que des flux de gènes importants ont lieu à un niveau régional.

67 Chapitre 2

68 Chapitre 3

. Suivi d;une épidémie dans un système en corridor et interactions entre compartiments ' sauvage et cultivé Epidémie en corridor

Les flux de gènes entre les populations d'agents pathogènes évoluant au sein de compartiments de plantes hôtes cultivées et sauvages, respectivement, peuvent revêtir une grande importance dans la gestion des résistances aux maladies. En effet, les populations pathogènes hébergées dans le compartiment sauvage constituent un réservoir de diversité génétique à partir duquel peuvent émerger des individus qui seront capables de contourner les gènes de résistance déployés dans le compartiment cultivé. De façon complémentaire, l'introgression dans le compartiment sauvage de facteurs de virulence préalablement sélectionnés dans le compartiment cultivé peut jouer le rôle de refuge pour ces mêmes virulences. De nombreux gènes de résistance complète à M larici-populina ont été sélectionnés et diffusés dans des cultivars pour la lutte contre la rouille du peuplier. Toutes ces résistances ont à ce jour été contournées par l'agent pathogène (Pinon & Frey, 2005). Pour étudier les échanges génétiques entre les populations de M larici-populina des compartiments cultivé et sauvage, nous nous sommes intéressés à un système naturel qui présente l'avantage d'être assez isolé des grandes zones populicoles et relativement simplifié: la vallée de la Durance. Dans cette vallée encaissée, il existe une ripisylve naturelle à Populus nigra avec, dans la partie la plus en amont, une zone de sympatrie entre peupiier et mélèze, l'hôte alternant de M larici-populina qui permet la reproduction sexuée du champignon. De plus, de petites plantations de peupliers hybrides (cultivar 'Beaupré') possédant la résistance complète R7 ont été identifiées. Le but de cette étude est de comprendre comment se développe l'épidémie dans la vallée de la Durance et de déterminer si et comment les populations de M larici-populina des deux compartiments interagissent entre elles : les épidémies dans les deux compartiments se font­ elles séparément ou existe-t-il un brassage génétique entre populations pouvant mener à une introgression d'allèles de virulence dans le compartiment sauvage ou à une diversification accrue des populations du compartiment cultivé? Pour tenter de répondre à ces questions, des populations de M larici-populina situées le long du corridor de la Durance et caractéristiques des deux compartiments cultivé et sauvage ont été échantillonnées et les isolats analysées à l'aide de marqueurs microsatellites et de marqueurs de virulence.

Cette étude est présentée sous la forme d'une publication en préparation: Barrès B, Halkett F, Andrieux A, Pinon J, Frey P, Co-occurrence oftwo simultaneous poplar rust epidemics along a 200 km riparian Populus nigra corridor revealed by microsatellite and virulence markers.

69 Chapitre 3

Co-occurrence of two simultaneous poplar rust epidemics along a 200 km riparian Populus nigra corridor revealed by microsatellite and virulence markers.

Benoît BARRÈS, Fabien HALKETT, Axelle ANDRIEUX, Jean PINON and Pascal FREY

INRA Nancy, Equipe de Pathologie Forestière, UMR 1136 Interactions Arbres­ Microorganismes, IFR 110, 54280 Champenoux, France

Correspondence: P. Frey. Fax: +33 383 394069; E-mail: [email protected]

70 Epidémie en corridor

Introduction

Resistance to pests and pathogens is one ofthe most important selection criteria for cultivated plants. Host resistance can be based on quantitative trait loci or on major resistance genes that are responsible for partial and complete resistance, respectively. Partial resistances are thought to be more durable, whereas complete resistances are likely to be overcome by the pathogen, leading to phytosanitary problems (Jorge et al., 2005). The breakdown of major resistance genes can be the result of a single nucleotide mutation in an avirulence gene (Joosten et al., 1994). Alternatively it may also be the result of the multiplication of a genotype which was preexisting in a reservoir host plant (Wisler & Norris, 2005). Wild host populations could play the role of pathogen reservoir (Dinoor 1974). Furthermore, the existence of pathogen reservoirs in wild relatives of cultivated crops may threaten sorne disease management strategies, such as the periodical recycling of resistance genes. Therefore, an accurate assessment of the extent of gene flow between pathogen populations from their wild and cultivated hosts is crucial for a better management of the available resistances. Poplar cultivation in Europe is highly intensive and there have been many efforts by breeders during the past decades for developing rust resistant cultivars. However, the Eurasian poplar rust fungus, Melampsora larici-populina, successively broke down aU the complete resistances deployed so far in poplar cultivation, causing severe economic losses (Pinon & Frey, 2005). One of the most recent examples is that of Populus x interamericana 'Beaupré' which carry the R7 resistance gene. This cultivar was successfuUy grown between 1980 and 1994, and was one of the most popular cultivar in Northeastern Europe during this period. In 1994, breakdown of the R7 resistance was detected in Belgium, France and UK and virulent isolates spread aU over Western Europe in less than five years (Pinon et al., 1998; Lonsdale & Tabbush, 2002). In the same period, the sanitary state ofnatural riparian Populus nigra stands seemed to be stable (Gérard et al., 2006). Moreover, the pathotype composition of populations from the cultivated and the wild compartments was shown to be radicaUy different, with almost no complex pathotypes on wild P. nigra stands (Pinon & Frey, 1997; Gérard et al., 2006). Such differences could be either due (i) to isolation of populations from both wild and cultivated compartments or (ii) to an important cost of unnecessary virulences in the wild host populations (Gérard et al., 2006).

71 Chapitre 3

To study the interactions between M larici-populina populations from wild and cultivated poplar stands, the Durance River valley provides a simplified system which has many interests: (i) This valley is deeply embanked in the Alps, especially in its upstream part, and relatively isolated from the main French poplar cultivation areas. (ii) A nearly continuous riparian forest of wild P. nigra is growing aIl along the riverside, approximately 0.5 to 2 km wide and 200 km long (Figure 1). (iii) In the upstream part of the valley, European larch (Larix decidua), the aecial host of M larici-populina, occurs naturally, sometimes adjacent to P. nigra. (iv) Previous observations have shown that the poplar rust epidemic begins each spring in the poplar-Iarch sympatry zone and spreads downstream the river each summer (Gérard et al., 2006). (v) Five cultivated poplar stands of variable acreage, containing P. x interamericana 'Beaupré' were detected in the Durance River valley or in its vicinity (Figure 1). (vi) Larch trees are present adjacent to one of these 'Beaupré' stands (Pont-du-Fossé), a situation highly favorable for the completion of M larici-populina's sexual cycle. Vir7 isolates ofM larici-populina were detected on this 'Beaupré' stand since 1999. Altogether, these features make the Durance River valley an interesting model to study the annual poplar rust epidemic in a one-dimensional corridor system and to assess the interactions between M larici-populina populations from wild and cultivated compartments. Gene flow between the two stands ofpoplars is suspected, but has never been finely studied. To better understand and evaluate interactions between M larici-populina populations from the wild and cultivated stands, we investigated in this study the genetic diversity of this pathogen using a selected marker (vir7/avr7 phenotype) in combination with nine neutral microsatellite markers. If gene flow between the two types of populations is very important, resulting in a global mixing of populations, then we expect the neutral genetic differentiation to be low between cultivated and wild strands, the gene diversity to be as high in both compartment and an integration ofvirulence 7 in the populations collected from the wild host. At the opposite, ifthere are two independent epidemics in both types of stands, then we will expect to found an important variation in allele frequencies of neutral markers in both populations and a low frequency ofvirulence 7 the wild compartment.

Materials and rnethods

72 Epidémie en corridor

Sampling strategy

Prelles and Mirabeau were the two sites that delimited the studied transect of approximately 200 km in the Durance River valley (Figure 1). Along this transect, 10 sites with wild P. nigra stands were selected, separated by distances ranging from 5 to 32 km (Table 1). AIl the five cultivated poplar stands containing P. x interamericana 'Beaupré' were also selected for the study. At Mirabeau and Manosque, the two cultivated stands covered quite large areas (50 and 21 ha, respectively) and both contained 20% of cultivars harboring the 'R7' resistance, including 'Beaupré'. The three remaining cultivated stands (at Valernes, Manse and Pont-du­ Fossé) were much smaller, with only approximately 100, 20 and 35 'Beaupré' trees, respectively. Furthermore, the 'Beaupré' trees at Pont-du-Fossé were directly adjacent to larch trees. The date of appearance of the first rust symptoms, as weIl as the intensity of the infection was monitored for each site between July and October 2004 by visual observations every three weeks. Each site was sampled once in October 2004, at the end of the epidemic. The size of the area in which infected poplar leaves were collected at each site was approximately 0.5 to 2 ha. For natural riparian sites each sample was isolated from a distinct leaf collected on different P. nigra trees. Whenever possible, up to 50 trees were randomly selected per site. In hybrid poplar plantations, infected leaves were collected from fewer trees (20 to 30) because of the difficulty to sarnple infected leaves on adult hybrid poplars. The focal character of the infection at Valernes allowed us to collect 30 leaves from only five trees. The infection level of P. nigra trees at St-Clément, Châteauroux, La Brillanne and Mirabeau was very low, probably owing to dry microclimatic conditions. Therefore, only 10, 18, 22 and 5 infected leaves could be collected at these sites, respectively. The same limitation was encountered in the hybrid poplar plantation ofManosque where only 8 infected leaves could be collected. The collected leaves were stored in an ice box until arrivaI to the laboratory. One uredinium per leaf was randomly selected and monouredinial isolates were grown on leaf disks ofPopulus x euramericana 'Robusta', as previously described (Gérard et al., 2006).

Pathotype identification

The M larici-populina isolates were inoculated on a poplar differential set allowing the detection of eight virulences, as previously described (Gérard et al., 2006). Because most of the hybrid poplars carrying a defeated complete resistance identified in the Durance River

73 Chapitre 3

valley were carrying the R7 gene, only the information regarding the vir7/avr7 phenotype was retained in the present study. Percentages of vir7 isolates were computed on the complete dataset for each population. After pathotype identification one sporulating leaf disks of P. x euramericana 'Robusta' for each isolate was stored at -20°C in Eppendorf tubes until DNA extraction.

DNA analysis

DNA was extracted using DNeasy® 96 Plant Kit (Qiagen). We followed the Fresh Leaves protocol (DNeasy® 96 Plant Handbook, September 2002) except that the infected leaf disks were disrupted with one tungsten carbide bead, suspended into 200 III of extraction buffer, during 2 X 1 min instead of 2 X 1.5 min at 30 Hz. DNA was eluted in a final volume of 200 Ill. Eleven microsatellite loci were chosen among those developed by Barrès et al. (2006). PCR were performed individually in a PTC-200 Peltier thermal cycler (Ml Research) using conditions previously described (Barrès et al., 2006), except for locus IlMLP31 where the PCR mix was modified as follows: 15 ng template DNA, 2 III of 10X reaction buffer, 3 mM MgCh, 0.7 Ilg/IlÎ BSA (Sigma), 0.2 mM dNTP, 0.5 U Taq polymerase (Sigma) and 0.2 IlM forward and reverse primers in a 20 ilL final reaction volume. To allow size and dye muItiplexing, forward primers were labeled with three different dyes (Proligo), D2 for f.lMLP13, IlMLP22, f.lMLP27 and IlMLP37, D3 for IlMLP20, IlMLP28 and IlMLP30, and D4 for f.lMLP09, f.lMLP12, IlMLP31 and IlMLP36. PCR products were separated, sized, and analyzed on a CEQTM 8000 Genetic Analysis System (Beckman CouIter). In order to reduce the number of analyses, PCR products were pooled in two sets of loci. Set A was made up of IlMLP09, IlMLP13, IlMLP27, IlMLP30 and IlMLP36 loci with volumes of2, 3, 4, 4 and 4 Ill, respectively, in a 34 III final volume, whereas IlMLP12, IlMLP20, IlMLP22, IlMLP28, f.lMLP3l and IlMLP37 loci were put together within Set B, with volumes of 3,4, 7, 4, 2 and 8 f.ll, respectively, in a 61 III final volume. InternaI size standards of 400 and 600 pb (Beckman CouIter), labeled with Dl dye, were used to genotype Set A and Set B, respectively, in a mixture containing 30 III of Sample Loading Solution (SLS, Beckman CouIter), 0.5 III of internaI size standard and 1 III ofeach ofthe marker sets. When chromatograms were ofpoor quality, or when a locus failed to amplify, PCR of the entire set were performed again. Ifthe analysis failed again, the individual was considered as a missing data. It should be noted that f.lMLP22 is a mitochondrial microsatellite locus (Barrès et al., 2006). Each allele at this locus was therefore considered as a different mitotype.

74 Epidémie en corridor

Data analyses

Identical genotypes were identified using Gimlet version 1.3.3 (Valière, 2002) among the whole dataset. The insufficient power of molecular markers could lead to identify individuals with the same multilocus genotype which are not clonaI. In order to identify multilocus genotypes that are statistically overrepresented in populations assuming panmixia, and thus that could be considered as belonging to the same clonaI lineage, the method described by Halkett et al. (2005) was used. The probability of observing n times a multilocus genotype in a population was computed using MLGsim software (Stenberg et al., 2003). Then the program, using a Monte Carlo simulation method, determines the significant threshold for the probability values for each population, taking into account sample size and allele frequencies. The significance level was fixed to 0.05 in the present study. Hence, a clone-corrected dataset was built, keeping only one individual per identical multilocus genotypes that were considered as clones, at each site. This dataset was used in aIl the subsequent analyses.

Genotypic diversity (G) was computed, following Stoddart & Taylor (1988): G =

l/"jjfxCx/Nee)Z), wherelx is the number of genotypes observed x times in the sample, and Nec is the number of isolates after clone-correction. The relative genotypic diversity (G/Nec) and a t­ test for the significance of differences between genotypic diversities were computed (Chen et al., 1994). Genotypic linkage disequilibrium and deviation from Hardy-Weinberg equilibrium were computed using GENEPOP 3.4 (Raymond and Rousset, 1995) on the clone-corrected dataset. Significant levels were adjusted subsequently using the sequential Bonferroni correction method (Rice, 1989). Because of physical linkage ofloci 12 and 13 (Barrès et al., 2006), they were replaced by a chimeric new locus, named IlMLP38, reconstructed using PHASE (Stephens et al., 2001). In order to understand which parameter could affect population structure, assignment tests were performed at the individuallevel using STRUCTURE version 2.1 software (Pritchard et al., 2000). The number of alleles (Na), allelic richness (Ar) and gene diversity (He) were estimated on clone-corrected data for each population and for each pathotype group, using FSTAT 2.9.3.2 (Goudet, 1995). Permutation tests were carried out using FSTAT in order to test if allelic richness and gene diversity were significantly different between M larici-populina populations collected from wild vs. cultivated poplar stands. Number of permutations was set to 10000.

75 Chapitre 3

AlI the genotypic differentiations and exact tests were conducted using GENEPOP 3.4. Pairwise

FST were estimated using the method of Weir & Cockerham (1984). Estimation of genetic diversity parameters can be biased by a limited size of the sample. Therefore, populations with less than seven individuals were discarded from further analyses. Shared alIele distances were computed between the eleven populations of M larici-populina using nine nuclear microsatellites markers. Since three M larici-populina populations colIected from wild poplar stands (RCB, MON and VLR) contained a significant number ofvir7 isolates (Table 2), these populations were split into vir7 and avr7 subpopulations. Shared alIele distances were also computed for the resulting 14 populations. Two unrooted cladograms were built using unweighted pair group method with arithmetic averages (UPGMA) and TREEVIEW 1.6.6 software (Page 1996). Bootstrap values were calculated from 10000 replications over markers. A nested AMOVA was performed to partition the variance among the two pathotype groups (vir7 vs. avr7), among populations within pathotype groups, among individuals within populations and within individuals, using Arlequin version 3.1 software (Excoffier et al., 2005). The significance of variance component estimates was tested by 10000 random permutations. Mantel's tests (Mantel, 1967) were computed with 10000 bootstraps to test thesignificance of

Spearman rank correlation coefficient using GENEPOP. Since the rust epidemic spreads annualIy on a narrow corridor of riparian P. nigra stands a10ng the Durance River, the epidemic can be considered as one-dimensional. Thus geographical distances were not log­ transformed, as they shou1d be for two-dimensional epidemics. The hypothesis ofan isolation by distance was therefore tested by plotting FsJ(l-Fsr) against geographical distances (Rousset, 1997). Analyses were performed on populations defined on both their location and their vir7/avr7 phenotype. Three tests were performed, with vir7, avr7 and aIl populations, respectively.

Results

Genetic diversity

Characteristics of the 15 populations are given in Table 2. A total of 428 isolates have been analyzed and 263 different genotypes were identified using 9 microsatellite markers. AlI loci were found to be polymorphic. The number of alIeles ranged from 2 (for J..lMLP20 locus) to

76 Epidémie en corridor

15 (for flMLP38 locus). Four mitotypes were identified among isolates. Most ofthe identical multilocus genotypes were statistically overrepresented and thus were considered as belonging to the same clonallineage (Table 2). The relative genotypic diversity (êTlNcc) was high in aIl populations ranging from 0.714 for RCB population to the maximum of 1.000 in 9 populations (CHT, VLR, MEE, BR!, MIR, B-PDF, B-CDM, B-VLR and B-MAN). Linkage disequilibrium tests were significant (P

number of individuals of the smallest population (BRI population, Nee = 4). Allelic richness and gene diversity ranged from 1.67 to 2.42 and from 0.239 to 0.384, respectively. The highest values were found in PRE population whereas the lowest value of allelic richness and gene diversity were encountered in BRI and B-VLR populations, respectively. AIl loci were found to be polymorphic in populations from wild poplar stands whereas aIl loci but one were polymorphic in populations from cultivated stands. AlI four mitotypes were identified in populations from wild stands whereas oniy two mitotypes (C and D) were found in populations from cultivated stands. Furthermore, the mean allelic richness of populations was significantly higher (P<0.05, P-value after 10000 permutations) in the wild than in the

cultivated compartment (mean Ar = 2.20 and 1.84, respectively). Similarly, the mean gene diversity was significantly higher in the wild stands compared to the cultivated stands (mean He = 0.311 and 0.259, respectively, P

77 Chapitre 3

Genetic structure

Pairwise FST between populations ranged from 0 to 0.297 (Table 3). Most of the highest FST

values were obtained when considering the B-VLR population. Such high FST values were

probably a consequence of the low number of isolates (Nee = 5) in this population. Pairwise

FST between populations from wild stands were significant (P<0.05) in 17/45 (38%) of the

case. Out often FST values, only 4 (40%) were found significant (P<0.05) between cultivated populations. Pairwise FST between populations sampled on P. nigra and populations sampled on 'Beaupré' were significant in 39 of the 50 (78%) possible tests. Furthermore, most of the non-significant differentiations between populations from both compartments were probably due to small population sizes (e.g. BRI and MIR populations). Shared allele distances between populations were ranging from 0.001 between B-MAN and B-MIR to 0.086 between MEE and B-CDM populations. On the unrooted cladogram, populations from cultivated stands were strongly clustered, whereas the clustering of the populations from wild stands was weaker (Figure 2a). As already mentioned, the amount of vir7 isolates in the populations from wild stands was highest for RCB, MON and VLR populations (Table 2). Furthermore, an excess of homozygosity was detected in RCB, MON and VLR populations (FIS of 0.142, 0.209 and 0.295, respectively), which may be a signal of a subpopulation structure. Assignment tests performed on individuals revealed a trend to cluster individuals according to their vir7/avr7 phenotype; however no clear pattern was detected (results not shown) probably due to the insufficient resolution power of the microsatellite markers used. These results nonetheless prompted us to split these three populations into vir7 and avr7 subpopulations. A second cladogram was built on subpopulations defined on both their location and their vir7/avr7 phenotype. AlI vir7 subpopulations collected from wild P. nigra stands clearly clustered with populations collected from cultivated stands (Figure 2b).

Genetic structure o/the vir7/avr7pathotype groups

Since the differentiation observed between vir7 and avr7 populations appeared to be greater than between populations from wild and cultivated stands, it appears that the belonging to a pathotype group is an element which strongly affects the genetic structure of M larici­ populina populations. Therefore, all individuals were clustered in two pathotype groups according to their vir7/avr7 phenotype and genetic characteristics were computed for each

78 Epidémie en corridor

group (Table 4). Relative genotypic diversity was not significantly different between the two pathotype groups. The avr7 group exhibit the highest genetic diversity: allelic richness, gene diversity and the number of private alleles were higher in this group than in the vir7 group (Table 4). The analysis of molecular variance performed on 9 nuclear microsatellite markers showed that more than 82% ofthe variation was attributable to individuals. Approximately Il% ofthe variance was found among individuals within populations. The two remaining levels, among populations within pathotype groups and among pathotype groups, explained 1.5 and 5.0% of the variance, respectively (Table 5). Structure among individuals within populations was very strong (FIS = 0.121) and highly significant (P

was low (Fsc = 0.016) even if significant (P

relatively strong (FCT = 0.050) and highly significant (P

Discussion

Two elements could have structured the genetic diversity ofM larici-populina populations in the Durance River valley. The first one is the separation between the wild and the cultivated compartments. This hypothesis implies that isolation between M larici-populina populations collected from P. nigra and populations collected from 'Beaupré' exists and that epidemics in both compartment are developing independently of each other with no or only few introgression events and gene flow. This scenario is supported by genetic analysis which c1usters populations from the cultivated compartment (Figure 2a). But it is difficult to identify which barriers prevent gene flow between the two compartments since M larici-populina is a wind-dispersed pathogen and since poplar stands ofthe two compartments were very close or adjacent in sorne cases (Figure 1). The second hypothesis is the separation between vir7 and avr7 isolates, regardless the type of stand from which they were collected. This explanation was retained because it is coherent with the genetic structure observed and allows the understanding of the pattern observed between the two compartments. Populations from the cultivated compartment were sampled on 'Beaupré' which carries the R7 resistance gene and consequently aIl of the individuals of this compartment were vir7. Genetic c1ustering of vir7

79 Chapitre 3

populations from both the wild and cultivated compartments indicate a probable, even if not confirrned by an assignment test, 'cultivated' origin of vir7 individuals collected from the wild compartment. Therefore, exchange between compartments occurred, but in an unidirectional way: vir7 isolates from the cultivated stands could infect P. nigra, whereas avr7 isolates were not able to infect 'Beaupré'. However, could it be considered that M larici-populina populations from cultivated stands introgress populations from wild stands? Significant amounts ofvir7 isolates were only found in populations from the downstream part of the Durance River valley. At the opposite, in the most embanked part of the valley, between Prelles and Embrun, only few vir7 pathotypes (0­ 5%) were detected. These sites are located in the poplar-Iarch sympatry zone where sexual reproduction of M larici-populina occurs each spring, resulting in a true introgression of different isolates. Moreover, the isolation by distance pattern along the riverside was only detected between avr7 subpopulations from the wild compartment, and was not significant when aIl populations were considered, which is congruent with the hypothesis of a clonaI epidemic which spreads from the poplar-Iarch sympatry zone towards the downstream part of the valley. Altogether, our results suggest that, instead of an important introgression of M larici­ populina populations from the wild compartment with vir7 isolates from the cultivated compartment, there are two simultaneous epidemics, spreading annually downstream the Durance River valley. The first epidemic, mainly composed ofavr7 isolates, originates in the most upstream part of the valley, in the P. nigra - larch sympatry zone The second epidemic, mainly composed of vir7 isolates, may originate from the Pont-du-Fossé site, where P. x interamericana 'Beaupré' and larches are adjacent and spreads southwards, resulting in a mixture ofboth inocula on P. nigra southward ofRochebrune. To confirrn this pattern of migration, inter-annual surveys should be conducted. Because individuals that succeeded in migration along the river might be chosen at random, it could be expected that populations sampled in the downstream part of the valley would not be genetically stable over years. Since 2002, frequency of vir7 isolates at Prelles varied between oand 3% (data not shown). This is a clue that restricted introgression ofvir7 isolates in the M larici-populina populations from the wild poplar stands has already occurred. Two hypotheses can explain why vir7 frequency remains low in the wild compartment. The first explanation may be a cost ofthe virulence that affects the fitness of vir7 isolate. However, such a cost of unnecessary virulences has never been demonstrated in the Populus - M larici-populina pathosystem. Furthermore, the increase of pathotype complexity and the pyramiding of

80 Epidémie en corridor

unnecessary virulences observed in the past years in the cultivated stands does not support this hypothesis (Pinon & Frey 2005; Gérard et al., 2006). The second explanation may be a high resilience of the natural pathosystem, which could be due to a large effective M larici­ populina population size in the wild compartment and to the potential recessiveness of the virulence alle1e expected in a gene-for-gene system. These two features would result in a great genetic stability ofthe M larici-populina population which should slow down the integration ofvir7 alle1e in the wild compartment. Another interesting point is the lowest genetic diversity of vir7 populations compared to the avr7 populations. This could be due to a founder effect consecutive to the recent breakdown of the R7 resistance gene. Such a founder effect has already been observed in Venturia inaequalis populations after the breakdown ofthe apple scab resistance gene Vf(Guérin & Le Cam, 1994).

References

Barrès B, Dutech C, Andrieux A, Caron H, Pinon J, Frey P (2006) Isolation and characterization of 15 microsatellite loci in the poplar rust fungus, Melampsora larici­ populina, and cross-amplification in related species. Molecular Ecology Notes, 6, 60-64.

Dinoor A (1974) Role of wild and cultivated plants in the epidemiology of plant diseases in

Israel. Annual Review ofPhytopathology, 12,413-436.

Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online, 1,47-50.

Gérard P, Husson C, Pinon J, Frey P (2006) Comparison of genetic and virulence diversity of

Melampsora larici-populina populations on wild and cultivated poplar and influence of the altemate host. Phytopathology, 96, 1027-1036.

Goudet J (1995) FSTAT (version 1.2): A computer program to calculate F-statistics. Journal of Heredity, 86,485-486.

81 Chapitre 3

Guérin F, Le Cam B (2004) Breakdown of the scab resistance gene Vf in apple leads to a

founder effect in populations ofthe fungal pathogen Venturia inaequalis. Phytopathology, 94,

364-369.

Halkett F, Plantegenest M, Prunier-Leterme N, Mieuzet L, Delmotte F, Simon JC (2005) Admixed sexual and facultatively asexual aphid lineages at mating sites. Molecular Ecology, 14, 325-336.

Jorge V, Dowkiw A, Faivre-Rampant P, Bastien C (2005) Genetic architecture of qualitative and quantitative Melampsora larici-populina leaf rust resistance in hybrid poplar: genetic mapping and QTL detection. New Phytologist, 167, 113-127.

Lonsdale D, Tabbush P (2002) Poplar rust and its recent impact in Great Britain. Information

Note 7 (revised). Forestry Commission. Edimburgh, UK.

Mantel N (1967) The detection of disease clustering and a generalized regression approach. Cancer Research, 27, 209-220.

Nei M (1972) Genetic distance between populations. The American Naturalist, 106, 283-292.

Page RDM (1996) TREEVIEW: an application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences, 12, 357-358.

Pinon J, FreyP (1997) Structure of Melampsora larici-populina populations on wild and cultivated poplar. European Journal ofPlant Pathology, 103, 159-173.

Pinon J, Frey P, Husson C, Schipfer A (1998) Poplar rost (Melampsora larici-populina): the development of E4 pathotypes in France since 1994. Finnish Forest Research Institute,

Research Papers, 712, 57-64.

82 Epidémie en corridor

Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure usmg multilocus genotype data. Genetics, 155,945-959.

Raymond M, Rousset F (1995) GENEPOP (version1.2): population genetics software for exact tests and ecumenism. Journal ofHeredity, 86, 248-249.

Rice WR (1989) Analyzing tables ofstatistical tests. Evolution, 43, 223-225.

Rousset F (1997) Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Genetics, 145, 1219-1228.

Valière N (2002) GlMLET: a computer program for analysing genetic individual identification data. Molecular Ecology Notes, 2, 377-379.

Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358-1370.

Wisler GC, Norris RF (2005) Interactions between weeds and cultivated plants as related to

management ofplant pathogens. Weed Science, 53, 914-917.

Figure Legends

Figure 1: Location of the collection sites in the Durance River valley and in its vicinity (Southeastern France). Black squares correspond to hybrid poplar plantations and open squares correspond to wild P. nigra stands. The poplar-Iarch sympatry zone is indicated by an ellipse.

Figure 2: Unrooted UPGMA cladograms computed on DAS genetic distance using nine nuclear microsatellite markers between a) the Il populations of M larici-populina collected from the wild and the cultivated compartments and b) the 14 vir7/avr7 subpopulations. Populations with less than 5 individuals were discarded in both cases. Bootstrap values were computed on 10000 replicates. Since these values were lower than 50%, they were not indicated on the trees.

83 Chapitre 3

Figure 3: Fsrl(1-Fsr) plotted against the geographical distances, for aU pairwise comparisons between the seven avr7 M larici-populina subpopulations coUected from wild P. nigra stands. The P-value was obtained by a Mantel test (10000 permutations), computed with

GENEPOP program.

84 Epidémie en corridor

Tables and Figures

Figure 1: Barrès et al.

85 86 Epidémie en corridor

0.08

0.07

0.06

0.05

~ 0.04 '+ e 1;; r.. 0.03

0.02 y = O.OOOSx- 0.0016 R2 ~ 0.493 0.01 P = 0.002

0 20 40 60 80 100 -0.01 Geographical distance (km)

Figure 3: Barrès et al.

87 Chapitre 3

Table 1: Characteristics ofthe collection sites

Population ID Location Host Latitude Longitude Elevation

PRE Prelles Populus nigra 44°51'00.7" N 6°34'47.2" E 1156

CLE St-Clément Populus nigra 44°39'03.9" N 6°35'09.2" E 878

CHT Châteauroux Populus nigra 44°37'17.6" N 6°32'46.8" E 850

EMB Embrun Populus nigra 44°33'36.9" N 6°30'42.2" E 802

RCB Rochebrune Populus nigra 44°27'36.0" N 6°11'54.9" E 647

MON Monêtier Populus nigra 44°22'49.6" N 5°56'43.4" E 545

VLR Valernes Populus nigra 44°14'54.0" N 5°55'18.2" E 482

MEE Les Mées Populus nigra 44°02'08.2" N 5°57'54.1" E 392

BRI La Brillanne Populus nigra 43°55'21.0" N 5°53'59.8" E 344

MIR Mirabeau Populus nigra 43°41'26.3" N 5°40'22.2" E 233

P. x interamericana B-PDF Pont-du-Fossé 44°40'28.2" N 6°13'58.6" E 1200 'Beaupré'

P. x interamericana B-CDM Manse 44°37'11.5" N 6°08'04.3" E 1247 'Beaupré'

P. x interamericana B-VLR Valernes 44°14'57.7" N 5°55'20.4" E 482 'Beaupré'

P. x interamericana B-MAN Manosque 43°48'56.9" N 5°48'59.1" E 296 'Beaupré'

P. x interamericana B-MIR Mirabeau 43°42'31.4" N 5°42'56.1" E 247 'Beaupré'

88 Epidémie en corridor

Table 2: Characteristics ofthe M larici-populina populations.

Number ofisolates Population Numberof Numberof Relative genotypic Numberof after clone-correction Mean number of Allelic richness Gene diversity Percentage of ID isolates (N) genotypes (N;) polymorphie Mitotype diversity (Gi Nec) alleles (Na ± SD) (A r ± SD) (Nec) loci (He ± SD) vir7 (N" r71N)

PRE 46 43 40 0.878 8/9 4.78 ±2.64 2.42 ± 0.97 0.384 ± 0.226 A,B,C,D 2.2% (1/46)

CLE 9 9 8 0.818 7/9 2.89 ± 1.45 2.18± 1.01 0.313 ± 0.264 B,C,D 0.0% (0/9)

CHT 16 5 5 1.000 5/9 1.78±0.83 1.73 ± 0.76 0.242 ± 0.246 D 0.0% (0/16)

EMB 43 38 36 0.905 8/9 4.67 ± 3.16 2.39 ± 1.15 0.361 ± 0.261 C,D 4.7% (2/43)

RCB 42 35 29 0.714 9/9 4.22 ± 1.72 2.29 ± 1.02 0.337 ± 0.257 B,C,D 33.3% (14/42)

MON 46 39 38 0.951 8/9 4.44 ± 2.24 2.40 ± 1.08 0.371 ± 0.273 B,C,D 39.1 % (18/46)

VLR 44 33 33 1.000 9/9 4.00± 1.87 2.35 ± 0.93 0.382 ± 0.234 B,C,D 36.4% (16/44)

MEE 43 11 II 1.000 7/9 3.11 ± 1.90 2.28±1.27 0.345 ± 0.336 B,C,D 2.3% (1/43)

BRI 22 4 4 1.000 5/9 1.67 ± 0.71 1.67 ± 0.71 0.269 ± 0.285 C,D 18.2% (4/22)

MIR 5 5 5 1.000 6/9 2.56 ± 1.51 2.33 ± 1.28 0.333 ± 0.315 C,D 60.0% (3/5)

B-PDF 19 19 19 1.000 8/9 2.44±0.88 1.76 ± 0.63 0.248 ± 0.222 C,D 100% (19/19)

B-CDM 27 11 II 1.000 7/9 2.33 ± 1.32 1.88 ± 0.80 0.273 ± 0.228 C,D 100% (27/27)

B-VLR 28 5 5 1.000 6/9 1.78 ± 0.67 1.70 ± 0.61 0.239 ± 0.215 C,D 100% (28/28)

B-MAN 8 8 8 1.000 6/9 2.11 ± 0.93 1.85 ± 0.76 0.270 ± 0.250 C,D 100% (8/8)

B-MIR 30 23 22 0.920 7/9 3.00 ± 1.73 2.02 ± 0.97 0.294 ± 0.260 C,D 100% (30/30)

89 Chapitre 3

Table 3: Pairwise subpopulation Fsr, estimated with Weir and Cockerham's e(above diagonal) a and geographical distances in kilometers (be1ow diagonal). Comparisons between populations from the wild vs. cultivated compartment are boxed in the right hand corner.

Wild Cultivated

PRE CLE CHT EMB RCB MON VLR MEE BRI MIR B-PDF B-CDM B-VLR B-MAN B-MIR

Wild PRE 0.010 ns 0.022 ns 0.008 * 0.011 *** 0.021 *** 0.023 *** 0.060 *** 0.014 ns 0.037 ns 0.051 *** 0.072 ** 0.182 *** 0.058 ** 0.047 ***

CLE 22 0.052 ns 0.016 ns 0.011 * 0.033 * 0.014 ns 0.068 * 0.051 ns 0.030 ns 0.067 *** 0.097 *** 0.271 ** 0.056 * 0.043 *

CHT 26 5 0.041ns 0.034* O.013ns O.Ollns 0.117* 0.050ns 0.092ns 10.088*** 0.171** 0.202** 0.120* 0.083**

EMB 33 12 7 0.007 ** 0.018 ** 0.014 ** 0.023 * 0.007 ns 0.006 ns 0.070 *** 0.105 *** 0.181 *** 0.059 * 0.045 ***

RCB 53 37 33 27 0.005 ns 0.009 ** 0.044 * -0.011 ns omo ns 0.034 ** 0.070 ** 0.203 ** 0.033 ns 0.035 ***

MON 73 59 55 49 22 0.000 ns 0.033 * -0.035 ns 0.002 ns 0.032 *** 0.067 ** 0.132 ** 0.015 ns 0.021 **

VLR 85 69 65 58 32 15 0.028ns -0.024ns -0.022ns 0.031* 0.070** 0.125* O.Ollns 0.032***

MEE 103 84 80 73 51 38 24 0.028 ns -0.013 ns 0.133 *** 0.159 *** 0.205 *** 0.106 *** 0.087 ***

BRI 117 98 93 86 64 51 36 14 0.019 ns 0.013 ns 0.065 ns 0.131 ns 0.000 ns -0.003 ns

MIR 148 129 125 118 95 80 65 45 32 0.082 * 0.150 ** 0.205 ns 0.075 ns 0.066 ns

Cultivated B-PDF 34 28 26 26 24 40 54 74 88 118 0.016 ns 0.239 ** 0.006 ns 0.027 **

B-CDM 44 36 33 31 18 31 45 66 80 110 10 0.297*** 0.034ns 0.035*

B-VLR 85 69 65 58 32 15 0 24 36 65 53 45 0.226 ** 0.215 ***

B-MAN 130 111 107 100 78 64 49 27 14 18 101 93 49 -0.005 ns

B-MIR 145 126 121 114 92 77 62 42 28 4 115 107 62 14

a Unbiased estimate of the P-value of log-likelihood based exact test on genotypic distribution using a Markov chain method (dememorization=SOOO, batches=SO, iterations=2000) in GENEPOP version 3.4. Significance levels are indicated by stars (*, ** and *** for P

90 Epidémie en corridor

Table 4: Characteristics ofthe pathotype groups as defined by their vir7/avr7 phenotype.

'avrT 'virT

Number ofisolates (N) 258 170

Number ofisolates after clone 175 113 correction (Nec)

Number ofgenotypes (Ng) 157 103

Relative genotypic diversity (Ô/N,,) 0.792 0.837

Number ofpolymorphic loci 9/9 9/9

Allelic richness (Ar ± SD) 5.72 ± 3.30 4.11 ± 1.90

Gene diversity (He ± SD) 0.369 ± 0.245 0.303 ± 0.238

Mitotype A,B,C,D B,C,D

Number ofprivate alleles 21 3

FIs 0.135 0.119

i'leD U'riP NA..1'1CY 1 Biblicthèque des Sciences Rue du Jardin Botanique - CS 20148 5460~ VILLERS LES NANCY CEDEX 91 Chapitre 3

Table 5: AMOVA performed on the 14 vir7/avr7 subpopulations of M larici-populina. Pathotype groups are defined according to their avr7 or vir7 phenotype. Each pathotype group comprised seven populations with at least six individuals.

Variance Percentage of Degree offreedom Sum ofsquares F-statistics components variation

Among pathotype 22.9 0.082 5.0 F = 0.050 groups CT ***

Among populations within pathotype 12 31.7 0.025 1.5 Fsc = 0.016 *** groups

Among individuals 251 431.4 0.185 11.3 FIS 0.121 within populations = ***

Within individuals 265 357.5 1.349 82.2 FIT = 0.178 ***

92 Epidémie en corridor

L'étude des populations de M larici-populina dans la vallée de la Durance avait pour but de mieux appréhender les relations entre les compartiments sauvage et cultivé et de mieux comprendre le déroulement de l'épidémie le long du corridor de la vallée de la Durance. Les populations de M larici-populina qui se développent dans les compartiments cultivé et sauvage sont significativement différenciées (FST = 0.03). Des individus portant la virulence 7 (vir7) ont été identifiés dans les populations prélevées sur les hôtes sauvages. Lorsque les populations de l'agent pathogène sont subdivisées selon leur phénotype vir7/avr7, la différenciation génétique entre compartiment s'en trouve renforcée (FST = 0.05). De plus lorsqu'on calcule les distances génétiques entre les populations définies selon leur pathotype et leur origine géographique, les populations vir7 du compartiment sauvage se trouvent clairement regroupées avec les populations du compartiment cultivé. On peut donc émettre l'hypothèse qu'il existe deux épidémies se propageant simultanément dans la vallée de la Durance. La première épidémie, majoritairement composée d'isolats avr7, débute dans la zone de sympatrie P. nigra - mélèze dans la partie amont de la vallée et se propage vers l'aval. Cette dynamique spatio-temporelle est étayée par l'existence d'un patron d'isolement par la distance. La deuxième épidémie, majoritairement composée d'isolats vir7, débuterait dans un site de sympatrie 'Beaupré' - mélèze (vraisemblablement à Pont-du-Fossé). Puis elle se propagerait vers le sud, via le Col de Manse et rejoindrait la vallée de la Durance en aval d'Embrun. Ceci expliquerait la fréquence plus élevée des isolats vir7 sur les P. nigra dans la portion aval du transect de la vallée de la Durance. Nos résultats montrent clairement des échanges de souches de M larici-populina du compartiment cultivé vers les populations du compartiment sauvage. Ces échanges sont essentiellement unidirectionnels, étant donné l'incapacité des isolats avr7 à se développer sur des peupliers à résistance R7. En revanche, l'introgression des allèles (notamment la virulence 7) des isolats du compartiment cultivé dans la population du compartiment sauvage semble très faible, vraisemblablement en raison d'une très grande résilience du pathosystème naturel. Comme cette étude le montre, les capacités de migration de M larici-populina sont importantes. Dans cet exemple, nous avons détecté des migrations d'individus entre compartiments et un isolement par la distance sur environ 150 km. Toutefois, comme l'a montré le contournement de la résistance R7 suivie de la dissémination de la virulence 7 à une grande partie de l'Europe, l'étude de la structure des populations de M larici-populina à une plus grande échelle serait d'un grand d'intérêt.

93 Chapitre 3

94 Chapitre 4

Effet de la migration aux échelles continentale et intercontinentale sur la structuration génétique des populations de M. larici-populina Structures génétiques continentale et intercontinentale

Les capacités de dispersion de la plupart des Urédinales sont extrêmement grandes. Le plus souvent dispersés par le vent, certains agents de rouille peuvent migrer sur des centaines, voire des milliers de kilomètres (Nagarajan & Singh, 1990). M larici-populina est un agent de rouille hétéroïque et macrocyc1ique qui, lors de sa phase épidémique, produit d'énormes quantités d'urédospores dispersées par le vent (Dowkiw et al., 2003). La dissémination aléatoire d'un grand nombre de spores peut aboutir à des patrons d'isolement par la distance entre les populations, car les gradients de dispersion sont souvent très raides. Au cours du siècle dernier, de nombreux événements de dispersion à longue distance de M larici-populina ont été observés. Cependant, aucune étude de génétique n'a été effectuée sur ces populations afin de mieux comprendre le déroulement de ces événements. Afin d'étudier les capacités migratoires de M larici-populina, nous avons échantillonné huit populations européennes qui ont été génotypées à l'aide de marqueurs microsatellites. Deux populations récemment fondées en Islande et au Canada ont également été échantillonnées, afin de tenter de caractériser des événements de dispersion à longue distance. Un tel type de dispersion devrait aboutir à une forte différentiation par rapport aux populations sources. De plus, le nombre d'individus formant ces nouvelles populations devrait probablement être faible, et une baisse de la diversité génétique devrait donc être observée.

Cette étude est présentée sous la forme d'une publication en préparation: Barrès B, Andrieux A, Pinon J, Frey P, Genetic structure of the poplar rust fungus Melampsora larici-populina in Europe and recent founder effects in Iceland and Canada.

95 Chapitre 4

Genetic structure of the poplar rust fungus Melampsora larici-populina in Europe and recent founder effects in Iceland and Canada

Benoît BARRÈS, Axelle ANDRIEUX, Jean PINON and Pascal FREY

INRA Nancy, Equipe de Pathologie Forestière, UMR 1136 Interactions Arbres­ Microorganismes, IFR 110, 54280 Champenoux, France

Correspondence: P. Frey. Fax: +33383394069; E-mail: [email protected]

96 Structures génétiques continentale et intercontinentale

Introduction

Gene flow is a broad source of variation in plant pathogen populations. It leads to the foundation of new populations and can have a strong and rapid impact on population genetic structure. For rust fungi, which are obligate biotrophic pathogens, aerial dispersal is essential for survival because they are completely dependent on living host tissue. Wind dispersal is a passive type oftransport, compensated by the huge numbers of spores produced by rust fungi (Nagarajan & Singh, 1990). This dispersal strategy could result in isolation by distance (!BD) pattern, because of the rapid decrease of the probability of dispersal with distance to the source. Together with infected plant material movement, aerial dispersal could also lead to long distance dispersal (LDD) events, which have been reported for several rust fungi (Nagarajan & Singh, 1990). The Eurasian poplar rust fungus Melampsora larici-populina causes severe damages and economic losses in poplar cultivation. This heteroecious macrocyclic rust fungus, alternating on larches (Larix spp.), produces five spore stages during its life cycle, three of them (basidiospores, aeciospores and urediniospores) being wind-dispersed (Frey & Pinon, 2004). The high dispersal capacities of the asexual spores (urediniospores) are illustrated by the spread of a new virulence aIl over Western Europe in less than five years (Pinon & Frey, 2005). However, detecting gene flow based only on pathogenicity characters is problematic, since virulences are significantly influenced by selection (McDermott & McDonald, 1993). Neutral molecular markers usually offer better opportunities for measuring gene flow. Therefore, the first aim ofthis study was to characterize the genetic diversity and the structure ofM larici-populina populations at the European continental scale, using recently developed microsatellite markers (Barrès et al., 2006). Peculiar interest will be turned on the relation between geographical and genetic distances. Long distance dispersal (LDD) is defined as the transport of a pathogen in a viable form capable of causing infection over a distance of 1000 km or more (Nagarajan & Singh, 1990). M larici-populina is native to Eurasia and its distribution is supposed to encompass that of its natural host, Populus nigra. Several documented LDD events during the past century have resulted in a widespread distribution, covering most of the poplar-producing regions worldwide. Outbreaks of M larici-populina on other continents were reported from South America (Spegazzini, 1918; Fresa, 1936; Kern & Thurston, 1954), Southern Africa (Lloyd, 1971; Gibson & Waller, 1972), and Australia (Walker et al., 1974). During the year 1973, M larici-populina has been introduced in New-Zealand from Australia via trans-Tasman air-

97 Chapitre 4

currents over a 2000 km distance (Wilkinson & Spiers, 1976). In North America, M larici­ populina was first detected in 1991 in the US Pacific Northwest (Washington State and Oregon) (Newcombe & Chastagner, 1993) and subsequently in California (Pinon et al., 1994). Recently, M larici-populina was also detected in the North Central United States (Steimel et al., 2005) and in Eastern Canada in 2002 (Innes et al., 2004). In 1999, M larici­ populina was also discovered in Southern Iceland, infecting Populus trichocarpa clones originating from Alaska (H. Sverrisson, personal communication). Following its discovery on poplars, aecia of M larici-populina were found on larch needles (Larix spp.) in New Zealand (Wilkinson & Spiers, 1976), Canada (Grondin et al., 2005) and Iceland (H. Sverrisson, personal communication). It implies that M larici-populina had completed its life cycle, proceeding to sexual recombination, and has potentially durably settled in these new areas. Such colonization events, regardless the migration mechanism, have an important impact on the genetic structure of the newly established populations. Only a small fraction of individuals from the source population should have succeeded in LDD, resulting in a probable founder effect (Slatkin, 1977). Therefore, a low genetic diversity in founder populations and a high genetic differentiation between founder and source populations are expected. LDD are rare and highly stochastic events, which makes them difficult to study. The second aim ofthis study was to examine two recent LDD events which resulted in the foundation ofnew M larici-populina populations in Canada and lceland.

Materials and methods

Sampling strategy

Poplar leaves infected with M larici-populina were collected by us and collaborative researchers in 10 countries of the Northern hemisphere in the summer and autumn of 2003 (Figure 1, Table 1). Eight populations were collected from different European countries: Bosnia and Herzegovina (BIH), Czech Republic (CZE), France (FRA), Germany (DEU), Italy (ITA), the Netherlands (NLD), Poland (POL) and United Kingdom (GBR). The two remaining populations were collected from Iceland (ISL) and Canada (CAN), and were considered as recently founded populations. In each country, approximately 30-100 rust­ infected leaves were harvested from severa1 poplar trees over a total area ranging from 50 to

2 400 m • Six of the populations were collected from Populus nigra natural stands, whereas the Icelandic population was collected from P. trichocarpa, and the three remaining populations

98 Structures génétiques continentale et intercontinentale

from different hybrid poplars, mainly P. X euramericana and P. x interamericana (Table 1). The collected leaves were brought back or sent to the laboratory. One uredinium per leaf was randomly selected and grown on fresh leaf discs ofP. x euramericana 'Robusta' as described by Gérard et al. (2006). Approximately 2 mg ofurediniospores were collected for each isolate and stored at -20°C in Eppendorftubes until DNA extraction.

DNA analysis

DNA was extracted using DNeasy® 96 Plant Kit (Qiagen). We followed the Fresh Leaves protocol (DNeasy® 96 Plant Handbook, September 2002) except that samples were disrupted with one tungsten carbide bead, suspended into 200 /lI of extraction buffer, during 2 X 1 min instead of 2 X 1.5 min at 30 Hz. DNA was eluted in a final volume of 200 /lI. Eleven microsatellite loci were chosen among those developed by Barrès et al. (2006). PCR were performed individually in a PTC-200 Peltier thermal cycler (Ml Research) using conditions previously described (Barrès et al., 2006), except for locus /lMLP31 where the PCR mix was modified as follows: 15 ng template DNA, 2 /lI of 10X reaction buffer, 3 mM MgCh, 0.7 /lgl/ll BSA (Sigma), 0.2 mM dNTP, 0.5 U Taq polymerase (Sigma) and 0.2 /lM forward and reverse primers in a 20 IlL final reaction volume. To allow size and dye multiplexing, forward primers were labeled with three different dyes (Proligo), D2 for /lMLP13, /lMLP22, /lMLP27 and /lMLP37, D3 for /lMLP20, /lMLP28 and /lMLP30, and D4 for /lMLP09, /lMLP12, /lMLP31 and /lMLP36. PCR products were separated, sized, and analyzed on a CEQTM 8000 Genetic Analysis System (Beckman Coulter). In order to reduce the number of analyses, PCR products were pooled in two sets ofloci. Set A was made up of /lMLP09, /lMLP13, /lMLP27, /lMLP30 and /lMLP36 loci with volumes of 2, 3, 4, 4 and 4 /lI, respectively, in a 34 /lI final volume, whereas /lMLP12, /lMLP20, /lMLP22, /lMLP28, /lMLP31 and /lMLP37 loci were put together within Set B, with volumes of 3, 4, 7, 4, 2 and 8 /lI, respectively, in a 61 /lI final volume. Internal size standards of 400 and 600 pb (Beckman Coulter), labeled with Dl dye, were used to genotype Set A and Set B, respectively, in a mixture containing 30 /lI of Sample Loading Solution (SLS, Beckman Coulter), 0.5 /lI of internaI size standard and 1 /lI ofeach of the marker sets. When chromatograms were ofpoor quality, or when a locus failed to amplify, PCR of the entire set were performed again. If the analysis failed again, the individual was considered as a missing data. It should be noted that /lMLP22 is a mitochondrial microsatellite locus (Barrès et al., 2006). Each allele at this locus was therefore considered as a different mitotype.

99 Chavitre 4

Data analyses

Identical genotypes were identified using Gimlet version 1.3.3 (Valière, 2002) for pooled populations and for each site. The insufficient power of molecular markers could lead to identify individuals with the same multilocus genotype which are not clonaI. In order to identify multilocus genotypes that are statistically overrepresented assuming panmixia, and thus that could be considered as belonging to the same clonaI lineage, the method described by Halkett et al. (2005) was used. The probability ofobserving n times a multilocus genotype in a population was computed using MLGsim software (Stenberg et al., 2003). Then the program, using a Monte Carlo simulation method, determines the significance threshold for the probability values for each population, taking into account sample size and allele frequencies. The significance level was fixed to 0.01 in the present study. Hence, a clone­ corrected dataset was built, keeping only one individual per identical multilocus genotypes that were considered as clones, at each site. Genotypic diversity (Ô) was computed, following

Stoddart & Taylor (1988): Ô = lri/JXx/NeeY) , where.lx is the number of genotypes observed x times in the sample, and Nec is the number of isolates after clone-correction. The relative genotypic diversity (Ô/Nec) and a t-test for the significance of differences between genotypic diversities (Chen et al., 1994) were computed. Genotypic linkage disequilibrium and deviation from Hardy-Weinberg equilibrium were computed using GENEPOP 3.4 (Raymond and Rousset, 1995) on the clone-corrected dataset. Significant levels were adjusted subsequently using the sequential Bonferroni correction method (Rice, 1989). Because of physical linkage of loci /-lMLP 12 and /-lMLP 13 (Barrès et al., 2006), they were replaced by a chimeric new locus, named /-lMLP38, reconstructed using PHASE (Stephens et al., 2001). Number of alleles (Na), allelic richness (Ar), gene diversity (He) and inbreeding coefficient (FIS) were estimated on clone-corrected data for each population using FSTAT 2.9.3.2 (Goudet, 1995). A test to detect recent founder effects was conducted using BOTTLENECK version 1.2 (Piry et al., 1999) on the Canadian and the Icelandic populations. This test is based on the assumption that populations that have experienced a recent reduction oftheir effective size exhibit a faster reduction of their allele number than of their gene diversity. The program computes the Wilcoxon's test for gene diversity excess after estimating the expected gene diversity at mutation-drift equilibrium using three mutation models: infinite allele model, stepwise mutation model and two-phase modeI. Permutation tests were also carried out using FSTAT in

100 Structures génétiques continentale et intercontinentale

order to test if aIlelic richness and gene diversity were significantly different between M larici-populina populations coIlected from their native distribution area (i. e. European populations) vs. recently founded populations (i.e. Canadian and Icelandic populations), and also between European populations sampled on hybrid poplars (FRA, GBR) vs. European populations sampled on P. nigra (BIH, CZE, DEU, ITA, NLD, POL).

AlI the genotypic differentiations and exact tests were conducted using GENEPOP 3.4. Pairwise

FST were estimated using the method of Weir & Cockerham (1984). Nei's minimum genetic distance (1972) was computed between the ten populations ofM larici-populina using the 9 nuc1ear microsateIlite markers. An unrooted tree was built using unweighted pair group method with arithrnetic averages (UPGMA) and DRAWGRAM software from the PHYLIP 3.63 package (Felsenstein 1989). Bootstrap values were ca1culated from 1000 replications over markers. Spatial analyses were perforrned usmg two methods. First, an autocorrelation analysis developed by Smouse and PeakaIl (1999) for multiaIlelic codominant loci was computed. Such an analysis is intrinsicaIly multivariate and thus strengthens the spatial signal by reducing stochastic noise. Spatial correlograrns were generated using GenAlEx 6 (PeakaIl & Smouse, 2005) and showed the genetic correlation as a function of distances. The 95% confidence interval about the autocorrelation coefficient r was estimated via 1000 bootstrappings. Tests for statistical significance were also performed by 1000 random permutations, estimating the 95% confidence interval of r about the nuIl hypothesis of the spatial genetic structure (Smouse & Peakall, 1999). The distance class size was set to 500 km. Second, correlations between genetic and geographic distances were complemented using Mantel correlation tests (Mantel, 1967). Two datasets were used: the first contained aIl the studied populations and the second excluded the Canadian and Icelandic populations.

Significance of Spearman rank correlation coefficients was computed using GENEPOP. The hypothesis of an isolation by distance was tested on the clone-corrected dataset by plotting

FsrI(1-FsT) against log-transformed geographical distances (Rousset 1997).

ResuUs

Genetic diversity

In aIl, 177 M larici-populina isolates were included in this study. Among these, 101 distinct genotypes were identified. No identical genotype was detected in two different populations.

101 Chapitre 4

The mean expected heterozygosity over aU loci and aU sites was moderately high (He = 0.346 ± 0.265). AU loci were found to be polymorphic and exhibited numbers of aUeles ranging from 2 (for JlMLP20 and JlMLP37 loci) to 10 (for JlMLP38 locus). Within European populations, most ofidentical multilocus genotypes were considered as non­ clonaI according to the analysis performed with MLGsim, except in the population from Bosnia and Herzegovina where 21 individuals were found to result from asexual reproduction (Table 1). Thus the clone-corrected dataset was used in further analyses. Out of the 222 possible linkage disequilibrium tests, only 15 were found to be significant (P<0.05). None of the tests remained significant when using the adjusted method level of significance. None of the European populations was found to deviate significantly from Hardy-Weinberg proportions. In the European populations, the number of polymorphic loci varied among populations from five, in GBR, to nine in DEU or in ITA. The relative genotypic diversity was high in aU but GBR population and ranged from 0.429 to 1.000 (Table 2). AUelic richness was moderately high (Ar = 2.66 ± 0.38, mean ± SD), ranging from 1.78 to 3.05, in GBR and ITA populations, respectively. In the European populations, the highest and lowest values of gene diversity were also found in ITA (He = 0.431) and in GBR (He = 0.236) populations, respectively, and were globally high (He = 0.351 ± 0.064, mean ± SD). Mean aUelic richness and mean gene diversity were not significantly different (P = 0.31 and P = 0.54, respectively) between populations coUected either on P. nigra or on hybrid poplars. M larici-populina populations from Canada and Iceland are known to be recently founded (Innes et al., 2004; H. Sverrisson, personal communication). None of the 13 linkage disequilibrium tests were found significant in CAN and ISL populations, consistently with the occurrence of sexual reproduction on larches in both populations. Whereas European populations seemed to respect Hardy-Weinberg proportions, this was not the case for CAN and ISL populations. A highly significant (P

Canadian and Icelandic populations, respectively. Mean aUelic richness (Ar = 1.43 ± 0.19, mean ± SD) and mean gene diversity (He = 0.131 ± 0.091, mean ± SD) were significantly lower (P<0.05) in these two recent populations compared to European populations. Furthermore, aU the aUeles found in CAN and ISL populations were found in European populations. Two to four mitotypes were found in Europeans populations whereas only one was identified in CAN and ISL populations (Table 2). Therefore, genetic diversity of Canadian and Icelandic populations represents only a subset of the existing diversity in the

102 Structures génétiques continentale et intercontinentale

European populations. Independently of the mutation model assumed, the tests performed with BüTTLENECK were not significant for both CAN and ISL populations.

Genetic dijferentiation

Nei's minimum genetic distances between European populations were rather low to moderate. Populations originating from Europe, and more especiaUy those from CZE, POL, FRA, NLD and DEU, were found to be clustered, whereas CAN and ISL populations show great genetic distances with European populations (Figure 2). Significant genetic differentiation was detected among European populations (FST = 0.057, P<0.001). Pairvvise FST between populations ranged from 0 to 0.162 (Table 3). Greatest differentiations were found between the GBR and populations from continental Europe. Pairwise FST between European and the two recently founded populations were aU very important and highly significant (P<0.001) ranging from 0.250 to 0.653 (Table 3). Furthermore, the genetic differentiation between CAN and ISL populations was also very high (FST = 0.653, P<0.001).

Spatial genetic structure

Spatial genetic autocorrelation analysis was computed for a distance class size of 500 km (Figure 3). The autocorrelation index r yielded positive and significant values at 500, 1000 and 1500 km, with an x-intercept at 1915 km, and then became significantly negative at 2500 km (Figure 3). After this decline, the correlogram showed oscillations of negative autocorrelation values, which is in accordance with computed simulations under restricted gene flow (Epperson et al., 1999; Turner et al., 1982). The first x-intercept value provides an estimate of the extent of non random genetic structure (Peakall et al., 2003). Significant negative r vaIues were detected between 3500 and 5500 km. This pattern was probably an effect of populations resulting from LDD events which were not expected to exhibit a correlation between genetic and geographical distances. A Mantel test was performed on aIl the populations, including CAN and ISL. The same test was conducted on European populations only, because CAN and ISL populations were not expected to show an isolation by distance pattern. Indeed, the estimate of the extent of non random genetic structure obtained with the autocorrelation analysis (i.e. 1915 km) was lower than the mean distance between European populations and CAN and ISL ones. Significantly (P<0.01) positive correlations between Fs/(l-Fst) and the log-transformed geographical

103 Chapitre 4 distances for both tests were detected (Figure 4). These results were broadly consistent with the autocorrelation results and revealed that proximate M larici-populina individuals were genetically more alike than more distant ones.

Discussion

Effect oflong distance dispersal on genetic diversity

Introduction of M larici-populina in Eastern Canada and Iceland are recent events (Innes et al., 2004; H. Sverrisson, personal communication). Such long distance dispersal events should have been caused by a limited number of individuals. As a consequence, strong founder effects should be observed. If this is the case, effects on genetic structure and diversity are expected from theoretical models. First, the genetic diversity should be reduced compared to populations from the center of origin. Indeed, in the Canadian and Icelandic populations, most ofthe loci were found monomorphic and allelic richness and gene diversity were significantly lower than in European populations. Second, founder effects may result in an important genetic differentiation among populations because of the rapid clonaI amplification of few individuals resulting in a rapid divergence of gene frequencies (Boileau et al., 1992). Indeed, a high genetic structuration was detected among the two newly founded populations and between these populations and European ones. These findings strongly suggest the independence of the two introductions and illustrate the stochastic character of these events. Cornuet & Luikart (1996) showed that recent founder effects result in a gene diversity excess at selectively neutral loci. The tests performed with BüTTLENECK on CAN and ISL populations were not significant, which may be explained by the low number of polymorphie loci in these populations resulting in a lack of power of the tests. Canadian and lcelandic populations seemed not to have reached the Hardy-Weinberg equilibrium, which is consistent with a recent introduction of the pathogen. In Canada, it seems that the M larici-populina population was founded by a very restricted nurnber of individuals, resulting in a high rate of selfing (F~s = 0.771, Table 2). ln lceland, the occurrence ofnull alleles was suspected because amplification of sorne loci failed for several isolates (data not shown). This could also lead to deviation from Hardy-Weinberg proportions. Long distance dispersal events resulting in founder effects were already reported for other rust fungi such as coffee leaf rust (Hemileia vastatrix) (Bowden et al., 1971) and sugarcane rust (Puccinia melanocephala) (Purdy et al., 1985), which both are supposed to have been

104 Structures génétiques continentale et intercontinentale

transported by wind across the Atlantic Ocean. Recent surveys of Mycosphaerella jijiensis populations demonstrated the recent stochastic spread of this banana pathogen at the intercontinental and continental scales (Carlier et al., 1996; Rivas et al., 2004). Founder effects were associated with limited airborne ascospores dispersal or movement of infected plant material. The development of M larici-populina epidemics in Iceland and Canada were probably subsequent to the introduction of a small number of urediniospores into these countries. Long distance wind-dispersal could explain the migration of spores from Europe to lceland. A similar LDD event has already been reported for M larici-populina between Australia and New-Zealand via trans-Tasman air-currents (Wilkinson & Spiers, 1976), but also for other rust fungi. Furthermore, insects are often transported by wind from Europe to Iceland, especially Lepidopteran species. In total 118 Lepidopteran species have been recorded in Iceland, thereof are 27 immigrant species (Olafsson & Bj5rnsson, 1997). Large amount ofbirch pollen originating from continental Europe have also been detected in Iceland in May 2006, several weeks before the flowering oflocal birch trees (M. Hallsdottir, personal communication). The hypothesis of introduction of M larici-populina through infected plant material was discarded, since no poplar cuttings were imported from Europe to Iceland (H. Sverrisson, personal communication). Why poplars remained uninfected by any rust fungus during decades in Iceland probably refiect the very low probability of viable spores to reach susceptible poplar leaves. However, several factors may have increased this probability in recent years. The development of poplar cultivation areas in Iceland during the 1990's increased the net trapping effect, and severe poplar rust epidemics in Europe during the period 1996-2000 certainly strengthened the inoculum pressure (Lonsdale & Tabbush, 2002; Pinon & Frey, 2005). The wind dispersal hypothesis seems less realistic to explain the introduction of M larici­ populina in Canada, considering (i) the higher distance and (ii) the direction of the prevailing winds in the Northern hemisphere. Moreover, most of the Populus species from North America are susceptible to M larici-populina. Therefore, if wind dispersal migration could have occurred in the past, M larici-populina should have been discovered far before its first report in North America (Newcombe & Chastagner, 1993). Wilkinson and Spiers (1976) suggested the possible spread of M larici-populina by infected plant material to explain the introduction of the pathogen in Australia. However, no evidence of survival of M larici­ populina as urediniospores attached to poplar buds or as mycelium in buds was reported so far. The introduction of M larici-populina in Canada seems more likely to be due to human transport (e.g. by urediniospores carried on clothes).

105 Chapitre 4

It appears that M larici-populina has durably established in Canada and Iceland. Indeed, larches (Larix spp.) are present in both countries and aecia ofM larici-populina are observed each spring on larch needles in the vicinity of poplars in Canada (Grondin et al., 2005) and Iceland (H. Sverrisson, personal communication). The lack of significant linkage disequilibrium between microsatelIites loci is in agreement with the occurrence of sexual reproduction in the CAN and ISL populations. These findings underline the importance of pest alert networks and quarantine regulations in order to avoid the introduction of exotic pathogens in healthy areas, which can be established for many years if favorable conditions are gathered. Hybridization with aboriginal species is also a great danger of such introductions (Brasier, 2001). Hybridization between M larici-populina and M medusae f. sp. deltoidae, a North American poplar rust fungus, has already been reported in New Zealand (Spiers & Hopcroft, 1994) and South Africa (Frey et al., 2005). The hybrid taxon was shown to exhibit a cumulative host range ofboth parental species and could thus be a potential threat for poplar cultivation.

Isolation by distance

Genetic differentiation between the European M larici-populina populations was moderately high, indicating an important gene flow at this scale. AlI the populations were at the Hardy­ Weinberg equilibrium and none of the loci pairs were in linkage disequilibrium in any of the studied populations. This suggests that sexual reproduction occurs near aIl the locations where individuals were sampled. Indeed, European larch (Larix decidua) is native to the mountains of central Europe (Alps, Carpathians, Sudetes, Tatras) and lowlands in Northern Poland. Furthermore, larch has been widely planted in lowlands throughout Europe and is also widespread for ornamental purpose in public or private gardens. A positive spatial genetic structure has been detected when considering aIl the populations or only the European ones. However the extent ofnon random genetic structure was estimated to 1915 km, according to the autocorrelation analysis. Canadian and Icelandic populations are beyond this limit and as described earlier, are supposed to result from LDD events. Hence, studying the relationship between genetic and geographical distances at the European scale should provide a better estimate of the IBD pattern, whereas including overseas populations tends to bias the estimation of M larici-populina dispersal. Non-random spatial genetic patterns can result from selection or from restricted gene flow. The latter hypothesis was retained because signaIs of spatial genetic structure due to selection are only expected at

106 Structures génétiques continentale et intercontinentale

specific coding loci or linked loci. The GBR population exhibited a lower value of genetic diversity and was genetically more distant from other European populations which may be explained by its geographic isolation. Indeed, the GBR site was located in Northem Ireland, a province where poplar cultivation is very rare. Furthermore, the English Channel, the North Sea and the Irish Sea are as many barriers that could limit dispersal of M larici-populina spores from the continent. However, it should be noticed that this population was collected from hybrid poplar cultivars and it could not be excluded that the genetic differentiation was partIy due to selection by host. Nevertheless, the French population also was collected from hybrid poplars but exhibit a similar level of diversity and differentiation than other European populations collected from P. nigra. Furthermore, mean allelic richness and mean gene diversity were not significantly different for European populations collected on P. nigra vs. populations collected on hybrid poplars. The extent of non-random genetic structure ofM larici-populina populations as revealed by this study is consistent with reports from other wind dispersed pathogens. It was demonstrated that live spores of pathogenic fungi that cause powdery mildew and brown rust of barley and wheat in Europe were transported for at least 600 km (Hermansen et al., 1978). Long distance migration of Puccinia striiformis f. sp. tritici clones was detected in Northwestem Europe: specific pathogen clones were thought to be dispersed from United Kingdom to Denmark, Germany and France (Hovm0ller et al., 2002). M larici-populina dispersal capacities seem to be large, although IBD was found at the European continent scale. Combined with the widespread distribution of larch in Europe, which allows recombination and overwintering of the pathogen, it provides favorable conditions for the rapid expansion of adapted genotypes and resulting in devastating outbreaks in cultivation (Brown & Hovm0ller, 2002). Such outbreaks have already happened. Hybrid poplar cultivars (e.g. 'Beaupré', 'Boelare') carrying the R7 resistance to M larici-populina were widely planted in Europe until the end of the 1990's. The breakdown ofthis complete resistance gene was detected in 1994 in Belgium and France and virulent isolates spread all over France and Northem Europe in only a few years (Lonsdale & Tabbush, 2002; Pinon et al., 1998). Such an outbreak illustrates the ineffectiveness of a control strategy relying solely on complete resistance genes at the continental scale. Rational deployment of resistant cultivars in time and space is often advocated as a mean for future long-term disease control in annual crops (Hovm011er et al., 2002). Nevertheless, such strategies are more difficult to apply to poplar cultivation because of the perennial nature of trees. Development of more durable resistance should therefore be considered (Dowkiw et al., 2003). Another means to control poplar rust epidemics would be

107 Chapitre 4

the limitation or the eradication of the alternate host. An historic example of successful eradication is that of barberry (Berberis vulgaris), the alternate host of the wheat stem rust fungus (Puccinia graminis) in North America (Roelfs & Groth, 1980). Total eradication of Larix spp. in Europe is senseless, but a better management of larch cultivation could probably improve the sanitary situation ofEuropean poplar cultivation.

Acknowledgments

This work was supported by INRA and a fellowship from the Région Lorraine. We thank Dalibor Ballian (Sumarski fakultet u Sarajevu, Sarajevo, Bosnia and Herzegovina), Ivana Salkova (Silva Tarouca Research Institute for Landscape and Ornamental Gardening, Pruhonice, Czech Republic), Irmtraut Zaspel (Institute for Forest Genetics and Forest Tree Breeding, Waldsieversdorf, Germany), Lorenzo Vietto (ISP, Casale Monferrato, Italy), Gert Kranenborg (ALTERRA, Wageningen, The Netherlands), Tadeusz Tylkowski (Polish Academy of Sciences, Kornik, Poland), and Halldor Sverrisson (Ice1andic Forest Research, Mogilsa, Reykjavik, Iceland) for collecting and sending rust-infected poplar leaves. We thank also Alistair McCraken (DARD, Belfast, UK), Ma1com Dawson (DARD, Loughgall, UK), Pierre Périnet (MRNFP, Québec, Canada), Richard Hamelin (Canadian Forest Service, Québec, Canada), Pierre Munnier (SRPV, Amiens, France), Thierry Deville (Braine, France) and Jacques Cornu-Langy (La Quincy, France) for their help for collecting rust-infected poplar leaves. The authorization of importation of M larici-populina-infected poplar leaves from outside the EU was obtained from the French Plant Protection Service (Service de la Protection des Végétaux), Permit No. JDB/03/388.

References

Barrès B, Dutech C, Andrieux A, Caron H, Pinon J, Frey P (2006) Isolation and characterization of 15 microsatellite loci in the poplar rust fungus, Melampsora larici­ populina, and cross-amplification in related species. Molecular Ecology Notes, 6, 60-64.

Boileau MG, Hebert PDN, Schwartz SS (1992) Non-equilibrium gene frequency divergence: persistent founder effects in natural populations. Journal ofEvolutionary Biology, 5, 25-39.

108 Structures génétiques continentale et intercontinentale

Bowden J, Gregory PH, Johnson CG (1971) Possible wind transport ofcoffee leaf rust across the Atlantic Ocean. Nature, 229, 500.

Brasier CM (2001) Rapid evolution of introduced plant pathogens VIa interspecific hybridization. Bioscience, 51, 123-133.

Brown JKM, Hovm011er MS (2002) Aerial dispersal of pathogens on the global and continental scales and its impact on plant disease. Science, 297, 537-541.

Cariier J, Lebrun MH, Zapater MF, Dubois C, Mourichon X (1996) Genetic structure of the global population of banana leaf streak fungus, Mycosphaerellafrjiensis. Molecular Ecology, 5,499-510.

Chen RS, Boeger JM, McDonald BA (1994) Genetic stability in a population of a plant pathogenic fungus over time. Molecular Ecology, 3, 209-218.

Cornuet lM, Luikart G (1996) Description and power analysis oftwo tests for detecting recent population bottlenecks from allele frequency data. Genetics, 144,2001-2014.

Dowkiw A, Husson C, Frey P, Pinon J, Bastien C (2003) Partial resistance ta Melampsora larici-populina leaf rust in hybrid poplars: genetic variability in inoculated excised leaf disk bioassay and relationship with complete resistance. Phytopathology, 93, 421-427.

Epperson BK, Huang Z, Li TQ (1999) Measures of spatial structure in samples of genotypes for multiallelic loci. Genetica, 73, 251-261.

Felsenstein J (1989) PHYLIP - Phylogeny Inference Package (Version 3.2). Cladistics 5, 164­ 166.

Fresa R (1936) Argentine Republic: Melampsora larici-populina in the Delta of Parana. International Bulletin ofPlant Protection, 10, 145-146.

109 Chapitre 4

Frey P, Gérard P, Feau N, Husson C, Pinon J (2005) Variability and population biology of Melampsora rusts on poplars. In: Rust diseases of Willow and Poplar (eds. Pei MH, McCracken AR), pp 63-72. CAB International, Wallingford, DK.

Frey P, Pinon J (2004) La rouille du peuplier: un pathosystème modèle. Biofutur, 247, 28-32.

Gérard P, Husson C, Pinon J, Frey P (2006) Comparison of genetic and virulence diversity of Melampsora larici-populina populations on wild and cultivated poplar and influence of the alternate host. Phytopathology, 96, 1027-1036.

Gibson lAS, Waller JM (1972) Notes from ODA plant pathology liaison officers. Pest Articles and News Summaries, 18,336-337.

Goudet J (1995) FSTAT (version 1.2): A computer program to calculate F-statistics. Journal of Heredity, 86, 485-486.

Goudet J, Raymond M, de Meeûs T, Rousset F (1996) Testing differentiation in diploid populations. Genetics, 144, 1933-1940.

Grondin J, Bourassa M, Hamelin RC (2005) First report of the aecial state of Melampsora larici-populina on Larix spp. in North America. Plant Disease, 89, 1242.

Halkett F, Plantegenest M, Prunier-Leterme N, Mieuzet L, Delmotte F, Simon JC (2005) Admixed sexual and facultatively asexual aphid lineages at mating sites. Molecular Ecology, 14, 325-336.

Hermansen JE, Torp D, Prahm LP (1978) Studies oftransport oflive spores ofcereal mildew and rust fungus across the North Sea. Grana, 17, 41-46.

Hovm0ller MS, Justesen AF, Brown JKM (2002) Clonality and long distance migration of Puccinia striiformis f. sp. tritici in north-west Europe. Plant Pathology, 51, 24-32.

Innes L, Marchand M, Frey P, Bourassa M, Hamelin RC (2004) First report of Melampsora larici-populina on Populus spp. in Eastern North America. Plant Disease, 88, 85.

110 Structures génétiques continentale et intercontinentale

Kern FD, Thurston HW (1954) Additional species of Uredinales from Colombia. Mycologia, 46,354-357.

Lloyd HL (1971) New records of Rhodesian fungi. 1. Melampsora larici-populina Klebahn. Kirkia, 8, 1-2.

Lonsdale D, Tabbush P (2002) Poplar rust and its recent impact in Great Britain. Information Note 7 (revised). Forestry Commission. Edimburgh, UK.

Mantel N (1967) The detection of disease clustering and a generalized regression approach. Cancer Research, 27, 209-220.

McDermott lM, McDonald BA (1993) Gene flow in plant pathosystems. Annual Review of Phytopathology, 31, 353-373.

Nagarajan S, Singh DV (1990) Long-distance dispersal of rusts pathogens. Annual Review of Phytopathology, 28, 139-153.

Newcombe G, Chastagner GA (1993) First report of the Eurasian poplar leaf rust fungus, Melampsora larici-populina, in North-America. Plant Disease, 77, 532-535.

Nei M (1972) Genetic distance between populations. The American Naturalist, 106, 283-292.

Olafsson E, Bjomsson H (1997) Fidrildi â Islandi 1995. Fjolrit Nâtturufraedistofnunar No 32 [Lepidoptera in Iceland 1995. Report ofthe Icelandic Institute ofNatural History No 32], 138 pp.

Peakall R, Smouse PE (2005) GenAlEx 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecular Ecology Notes, 6, 288-295.

Peakall R, Ruibal M, Lindenmayer DB (2003) Spatial autocorrelation analysis offers new insights into gene flow in the Australian bush rat, Rattusfuscipes. Evolution, 57, 1182-1195.

111 Chapitre 4

Pinon J, Frey P (2005) Interactions between poplar clones and Melampsora populations and their implications for breeding for durable resistance. In: Rust diseases ofWillow and Poplar (eds. Pei MH, McCracken AR), pp. 139-154. CAB International, Wallingford, UK.

Pinon J, Frey P, Husson C, Schipfer A (1998) Poplar rust (Melampsora larici-populina): the development of E4 pathotypes in France since 1994. Finnish Forest Research Institute, Research Papers, 712, 57-64.

Pinon J, Newcombe G, Chastagner GA (1994) Identification of races of Melampsora larici­ populina, the Eurasian rust fungus, on Populus species in California and V/ashington. Plant Disease, 78, 101.

Piry S, Luikart G, Cornuet JM (1999) BOTTLENECK: a computer program for detecting recent reductions in the effective population size using allele frequency data. Journal ofHeredity, 90, 502-503.

Purdy LH, Krupa SV, Dean JL (1985) Introduction ofsugarcane rust into the Americas and its spread to Florida. Plant Disease, 69, 689-693.

Raymond M, Rousset F (1995) GENEPOP (version1.2): population genetics software for exact tests and ecumenism. Journal ofHeredity, 86, 248-249.

Rice WR (1989) Analyzing tables ofstatistical tests. Evolution, 43, 223-225.

Rivas GG, Zapater MF, Abadie C, Carlier J (2004) Founder effects and stochastic dispersal at the continental scale of the fungal pathogen of bananas Mycosphaerella fijiensis. Molecular Ecology, 13,471-482.

Rousset F (1997) Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Geneties, 145, 1219-1228.

Slatkin M (1977) Gene flow and genetic drift in a species subject to frequent local extinctions. Theoretical Population Biology, 12, 253-262.

112 Structures génétiques continentale et intercontinentale

Smouse PE, Peakall R (1999) Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity, 82, 561-573.

Spegazzini C (1918) Cuarta contribuci6n a la Micologia chilena. Revista Chilena Historia Natural, 22, 95-104.

Spiers AG, Hopcroft DH (1994) Comparative studies of the poplar rusts Melampsora medusae, M larici-populina and their interspecific hybrid M medusae-populina. Mycological Research, 98, 889-903.

Steimel L, Chen W, Harrington TC (2005) Development and characterization ofmicrosatellite markers for the poplar rust fungi Melampsora medusae and Melampsora larici-populina. Molecular Ecology Notes, 5, 484-486.

Stenberg P, Lundmark M, Saura A (2003) MLGsim: a program for detecting clones using a simulation approach. Molecular Ecology Notes, 3, 329-331.

Stephens M, Smith NJ, Donnelly P (2001) A new statistical method for haplotype reconstruction from population data. The American Journal ofHuman Genetics, 68, 978-989.

Stoddart JA, Taylor JF (1988) Genotypic diversity: estimation and prediction in samples. Genetics, 118, 705-711.

Turner ME, Stephens JC, Anderson WW (1982) Homozygosity and patch structure in plant populations as a result of nearest-neighbor pollination. Proceedings ofthe National Academy ofSciences ofthe USA, 79, 203-207.

Valière N (2002) GIMLET: a computer program for analysing genetic individual identification data. Molecular Ecology Notes, 2, 377-379.

Walker J, Hartigan D, Bertus AL (1974) Poplar Rusts in Australia with comments on potential conifer rusts. European Journal ofForest Pathology, 4, 100-118.

113 Chapitre 4

Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358-1370.

Wilkinson AG, Spiers AG (1976) Introduction ofthe poplar rusts Melampsora larici-populina and M medusae to New Zealand and their subsequent distribution. New Zealand Journal of Science, 19, 195-198.

Figure Legends

Figure 1: Origin ofthe 10 lvi larici-populina populations collected in 2003.

Figure 2: Unrooted UPGMA tree computed on Nei's mmlmum genetic distance (1972) between the ten populations of M larici-populina with the complete dataset using nine nuclear microsatellite markers. Bootstrap values computed on 1000 replicates are indicated.

Figure 3: Correlograms showing the spatial correlation indices r as a function of distance, for a distance class size of 500 km. 95% confidence interval about the null hypothesis of the random distribution of M larici-populina individuals (dotted line) and 95% confidence error bars about r as determined by bootstrapping.

Figure 4: Regression plot between log-transformed geographic distances and Fs/(1-Fst). p values were obtained by a Mantel test (10000 permutations). Open triangles represent pairs of one overseas (Canada or Iceland) and one European population, black circ1es represent pairs ofEuropean populations.

114 Structures génétiques continentale et intercontinentale

Tables and Figures

Figure 1: Barrès et al.

115 Chapitre 4

United Kingdom

Iceland

, Canada

Figure 2: Barrès et al.

116 Structures génétiques continentale et intercontinentale

0.500

0.400

0.300

0.200

0.100 r .- .... 0.000

-0.100

-0.200

-0.300

-0.400 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 Geographical distance (km)

Figure 3: Barrès et al.

117 Chapitre 4

1.8

1.3

Il Y = 0.38x - 2.40 R2 =0.59 P=0.002

Fst/(1-Fst) 0.8

y = O.07x - 0.38 0.3 R2 = 0.52 e. .. P=0.006 ...... •• _0" • • 5.5 7 7.5 8 8.5 9 9.5

-0.2 In(distance(km»

Figure 4: Barrès et al.

118 Structures zénéti(Jues continentale et intercontinentale

Table 1: Characteristics ofthe collection sites ofM larici-populina populations

Country ISO code Location and province Host Latitude Longitude Collector

Bosnia and Herzegovina BIH Ovcarevo, Konjic Populus nigra 43°41'N 17°59' E D. Ballian

Czech Republic CZE Veltruby, Kolin Populus nigra 50°05' N 15°10' E I. Salkova

France FRA La Quincy, Picardie Hybrîd poplars 49°25' N 03°26' E P. Frey

Gennany DEU Ziltendorfer Niederung, Brandenburg Populus nigra 52°12'N 14°40' E I. Zaspe1

Italy ITA Zerbolo, Pavia Populus nigra 45°12' N 09°03' E L. Vietto

The Netherlands NLD Gendt, Gelderland Populus nigra 51°51' N 05°58' E G. Kranenborg

Poland POL Warta River, Poznan Populus nigra 52°12' N 16°53'E T. Tylkowski

United Kingdom GBR CastIearchdale, Northem Ireland Hybrîd poplars 54°28' N 07°43'W P. Frey

P. Frey Canada CAN Lotbinière, Québec Hybrîd poplars 46°30' N 71°55' W

20°31' W H. Sverrisson Iceland ISL Skalholt, Amessysla Populus trichocarpa 64°07' N

119 Chavitre 4

Table 2: Characteristics ofthe M. larici-populina populations

Numberof Relative Numberof Mean number of Numberof isolates after Numberof Genotypic Allelic richness Gene diversity Identifiant genotypic polymorphic alleles (Na ± FIS Mitotype isolates (N) clone-correction genotypes (N ) diversity (6) (A,± SD) (He± SD) g diversity (G/Nec) loci SD) (Nec)

BIH 31 10 9 8.3 0.833 8/9 2.78 ± 1.48 2.71 ± 1.41 0.401 ± 0.238 -0.137 B,C,D,E

CZE 10 9 7 6.2 0.692 8/9 2.67 ± 1.00 2.67 ± 1.00 0.321 ± 0.234 0.115 C,D

FRA 27 27 24 20.8 0.771 8/9 3.67 ± 1.87 2.74± 1.30 0.298 ± 0.240 -0.020 B,C,D

DEU 12 10 9 8.3 0.833 9/9 2.89 ± 1.36 2.82 ± 1.3 1 0.401 ± 0.239 0.059 C,D

!TA 12 10 9 8.3 0.833 9/9 3.11 ± 1.36 3.05 ± 1.34 0.431 ± 0.258 0.073 C,D,E

NLD 12 II 10 9.3 0.846 8/9 2.78 ± 1.09 2.65 ± 1.03 0.349 ± 0.238 0.104 C,D

2.83 ± 1.40 0.373 ± 0.237 0.225 B,C,D POL 10 10 10 10.0 1.000 8/9 2.89 ± 1.45

1.78 ± 0.83 1.78 ± 0.83 0.236 ± 0.270 -0.203 C,D GBR Il 9 5 3.9 0.429 5/9

1.44 ± 0.73 1.30 ± 0.48 0.067 ± 0.121 0.771 D CAè'I 30 29 6 2.3 0.080 3/9

5/9 1.78 ± 0.83 1.57 ± 0.61 0.196 ± 0.244 -0.007 D ISL 22 22 12 9.3 0.423

120 Structures génétiques continentale et intercontinentale

Table 3: Pairwise Fsr, estimated with Weir and Cockerham's 8 on the clone-corrected dataset between M larici-populina populationsa (above diagonal) and geographic distance between each location in kilometers (below diagonal).

Bosnia and Czech The United France Germany Italy Poland Canada Herzegovina Republic Netherlands Kingdom Iceland Bosnia and Herzegovina 0.056 ** 0.065 *** 0.029 ns 0.061 *** 0.040 ** 0.045 * 0.152 *** 0.474 *** 0.317 *** Czech 742 Republic 0.032 * 0.026 ns 0.072 * 0.027 ns -0.008 ns 0.141 ** 0.537 *** 0.297 ***

France 1256 850 0.047 ** 0.097 *** 0.015 ns 0.047 * 0.059 *** 0.488 *** 0.264 ***

Germany 981 240 851 0.053 *** 0.013 ns 0.001 ns 0.126 *** 0.469 *** 0.273 ***

Italy 698 715 633 885 0.062 ** 0.057 ** 0.162 *** 0.382 *** 0.250 ***

The 1258 677 324 600 774 Netherlands -0.007 ns 0.1 06 *** 0.551 *** 0.295 ***

Poland 961 267 999 153 973 753 0.147 *** 0.485 *** 0.260 ***

United 2187 1632 950 1509 1581 960 1654 Kingdom 0.653 *** 0.367 ***

Canada 7198 6621 5946 6459 6558 5967 6587 5014 0.653 ***

Iceland 3320 2635 2195 2434 2815 2074 2539 1322 4138

a Unbiased estimate of the P-value of log-likelihood based exact test on genotypic distribution using a Markov chain method (dememorization=SOOO, batches=SO, iterations=2000) in GENEPOP version 3.4. Significance levels are indicated by stars (*, ** and *** for P

121 Chapitre 4

L'étude moléculaire de huit populations européennes de M larici-populina a permis de montrer que, partout en Europe, les populations de cet agent pathogène présentent une diversité génétique importante. De plus, un patron d'isolement par la distance a été observé entre les populations du continent européen. Même s'il est difficile de donner une estimation précise de l'étendue géographique au sein de laquelle la structure génétique n'est pas due au hasard, celle-ci semble s'étendre sur près de 2000 km, d'après les analyses d'autocorrélation spatiale. Les populations d'Islande et du Canada étant situées à de grandes distances du continent européen, nous supposions que leur formation résultait d'événements de dispersion à longue distance. Ce phénomène étant de nature stochastique, une importante différentiation génétique entre les populations formées par ces évènements et les populations d'origine, ainsi qu'entre les populations formées entre elles, était attendue. Ces différentiations importantes ont été observées avec des Fsr atteignant des valeurs de plus de 60%. De plus, ces deux populations se sont révélées être extrêmement peu polymorphes et présentent toutes les deux une diversité génétique faible. Ces observations étayent l'hypothèse d'un effet de fondation consécutif à la migration à très longue distance d'un faible nombre d'individus. Il est difficile d'identifier les vecteurs de tels événements de dispersion, toutefois si l'hypothèse de l'anémochorie reste envisageable pour la population Islandaise, elle est très improbable dans le cas du Canada. Pour ce dernier événement, il est plus probable que l'homme en ait été le vecteur.

122 Conclusion & Perspectives Conclusion générale et perspectives

L'étude de la génétique des populations des champignons phytopathogènes permet l'inférence de paramètres biologiques qu'il serait par ailleurs extrêmement complexe d'évaluer de manière directe. Certains agents pathogènes, tels que les Urédinales, sont des microorganismes qui se multiplient rapidement, qui peuvent également combiner plusieurs modes de reproduction et qui ont des capacités de dispersion très étendue, parfois à l'échelle d'un continent. Toutes ces caractéristiques rendent difficiles, voire mêmes impossibles, l'évaluation de l'importance de la reproduction sexuée, la mesure des flux de gènes ainsi que l'estimation de la capacité de migration de telles espèces. Melampsora larici-populina fait partie de ces champignons pathogènes dont la diversité est difficile à étudier. Les connaissances sur la diversité génétique de M larici-populina étaient jusqu'ici assez limitées. Son cycle biologique très complexe, qui compte cinq types de spores, et se déroulant sur deux types d'hôtes, le peuplier et le mélèze, l'alternance d'une phase obligatoire de reproduction sexuée et de nombreux cycles de reproduction asexuée lors de la phase épidémique, ou encore le mode de dissémination par anémochorie, sont autant de facteurs qui peuvent influencer la diversité génétique de cet agent pathogène. Afin d'étudier cette diversité, des études ont été réalisées dans un premier temps à l'aide de marqueurs sélectionnés, les facteurs de virulence. Ces études ont permis de montrer que selon les pressions de sélection exercées par l'hôte, il existait une forte adaptation des populations de M larici-populina (Pinon & Frey, 1997 ; Pei et al., 2005). Ainsi la sélection de peupliers hybrides à résistance complète à la rouille a entraîné systématiquement le contournement de ces résistances complètes, provoquant parfois de graves épidémies et d'importantes pertes de rendement (Pinon & Frey, 2005). Les études à l'aide de marqueurs sélectionnés ont également permis de mettre en évidence un effet de la proximité du mélèze sur la richesse des populations en pathotypes (Frey et al., 2005). Cependant, l'étude de la diversité génétique à l'aide de marqueurs sélectionnés présente certains inconvénients. En effet, les pressions de sélection de l'hôte sur l'agent pathogène peuvent masquer l'influence des autres facteurs sur la diversité génétique. C'est pourquoi, le développement de marqueurs neutres a été entrepris. Les premiers marqueurs moléculaires développés sur M larici-populina furent des marqueurs RAPD (Random Amplified Polymorphism DNA) (Gérard et al., 2006) et AFLP (Amplified Fragment Length Polymorphism) (Pei et al., 2005). Ces marqueurs ont permis de mettre en évidence une très forte diversité génétique dans les populations de l'agent pathogène et des flux de gènes qui semblent important au niveau régional. Toutefois ces deux types de marqueurs génétiques

123 Conclusion générale et perspectives

sont dominants et sont donc mal adaptés à l'étude d'un organisme dicaryotique. Ils présentent de plus l'inconvénient d'être difficilement transférables entre laboratoires. C'est pourquoi nous avons décidé de développer des marqueurs microsatellites. Les marqueurs microsatellites sont codominants et réputés très polymorphes. En revanche, ils présentent l'inconvénient de devoir être isolés de nova pour chaque organisme étudié. Une stratégie d'enrichissement de l'ADN en motifs microsatellites a été utilisée avec succès pour développer 15 marqueurs microsatellites, dont 13 polymorphes, chez M larici-populina (Barrès et al., 2006). Grâce à l'isolement de ces marqueurs, nous avons abordé l'étude de la diversité génétique des populations de M larici-populina à plusieurs échelles spatiales. Trois études ont été effectuées. La première concerne l'organisation fine de la diversité. Pour cela un échantillonnage hiérarchisé et emboîté a été entrepris, constitué de quatre niveaux : la feuille, le rameau, l'arbre et le site. Cette première étude nous a permis de montrer que 90% de la diversité génétique chez M larici-populina était expliquée au niveau de la feuille. De plus, une structuration des populations en sous-populations à l'échelle de l'arbre a été montrée. La reproduction asexuée semble avoir un impact important à l'échelle du rameau sur la structuration des populations. Ces résultats nous ont permis de mieux cerner l'organisation spatiale des populations de M larici-populina et nous ont renseignés sur un mode d'échantillonnage pertinent de la diversité génétique des populations sur un site. Ainsi en pratique, pour éviter de prélever des individus résultant de la reproduction asexuée et disséminés à proximité directe, il est recommandé de n'échantillonner qu'un individu par rameau. La seconde étude a consisté à tenter de comprendre le déroulement d'une épidémie à l'échelle régionale, ainsi que l'interaction entre les populations de M larici-populina prélevées sur son hôte sauvage (P. nigra) et sur un hôte cultivé (Populus x interamericana 'Beaupré'). Dans le corridor de la vallée de la Durance, nous avons pu montrer qu'il existait des échanges entre les compartiments sauvage et cultivé, grâce au suivi des populations de M larici-populina à l'aide de marqueurs neutres et d'un marqueur sélectionné. Des isolats provenant de l'hôte cultivé ont été identifiés dans le compartiment sauvage. Il semble donc que l'introgression des allèles de virulence au sein du compartiment sauvage soit possible à terme. Ces hôtes sauvages sont donc également susceptibles de servir de zone refuge pour les agents pathogènes. De manière plus fine, nous avons également montré que l'épidémie dans le compartiment sauvage se développait de l'amont (la zone de sympatrie peuplier-mélèze) vers l'aval entraînant un isolement par la distance des populations le long du corridor de la

124 Conclusion générale et perspectives

vallée de la Durance. Ceci démontre l'aptitude de M larici-populina à une migration annuelle sur de grandes distances. C'est cette aptitude à la migration que nous avons voulu évaluer de manière plus générale et à une échelle plus large dans notre dernière étude. Dans le but d'évaluer les capacités maximales de dissémination de M larici-populina, nous avons effectué une étude de la structure génétique des populations de cet agent pathogène à l'échelle du continent européen. De plus, nous avons également étudié la diversité de deux populations récemment fondées au Canada et en Islande. Cette étude nous a permis d'identifier un patron d'isolement par la distance à l'échelle du continent européen. De plus nous avons également détecté un effet de fondation dans les populations résultant d'une dissémination à longue distance. Le caractère stochastique de tels événements a également été confirmé par l'étude de la différentiation génétique entre les populations.

La lutte chimique contre M larici-populina, même si elle est utilisée sur une faible proportion des peupleraies françaises (environ 15% des surfaces en Picardie; P. Munier, communication personnelle), est difficile à mettre en œuvre, en particulier lorsque les peupliers ont atteint une taille importante. De plus, elle se heurte à des obstacles financiers et environnementaux. Quelles sont alors les alternatives qui se présentent pour la lutte contre cet agent phytopathogène ? La gestion temporelle des cultivars, telle qu'elle est employée dans les cultures annuelles comme le blé ou l'orge dans certaines régions du monde, est tout à fait irréaliste dans le cas de la populiculture, car le peuplier est un arbre dont la culture dure au minimum une vingtaine d'années. D'après nos résultats (Chapitre 4), il semble que la gestion spatiale des différentes résistances complètes des cultivars de peuplier ne puisse être envisagée à une échelle autre qu'intercontinentale. Toutefois si une telle gestion était entreprise, les mesures de quarantaine devraient être renforcées pour éviter les événements de dispersion à longue distance qui ont été observés au cours du XXème siècle partout sur la planète, et qui sont probablement en grande partie le fait des échanges anthropiques. Ces événements peuvent aboutir à l'installation définitive de l'agent pathogène, si les conditions favorables sont réunies, comme nous avons pu l'observer au Canada ou en Islande. Les contournements successifs de toutes les résistances complètes sélectionnées et déployées à ce jour démontrent les grandes capacités d'adaptation et de migration de M larici-populina. La phase de reproduction sexuée, vraisemblablement obligatoire sous nos latitudes comme le montre l'étude des déséquilibres de liaison dans les populations, entraîne un brassage continuel du patrimoine génétique de M larici-populina. C'est pourquoi une stratégie de pyramidage des gènes de résistance complète chez le peuplier semble vouée à l'échec. On

125 Conclusion générale et perspectives

peut également souligner que de nombreux pathotypes très complexes ont déjà été observés dans la nature, dont celui possédant la totalité des huit virulences connues. Enfin, nous avons montré que les populations de M larici-populina du compartiment sauvage et du compartiment cultivé peuvent interagir et échanger des individus (Chapitre 3). Ainsi, on ne peut exclure à terme une introgression des nouveaux allèles de virulence dans le compartiment sauvage qui pourrait ensuite jouer le rôle de réservoir de virulences. L'amélioration variétale des peupliers cultivés, le développement de résistances durables, ainsi qu'une meilleure connaissance de la biologie et de l'épidémiologie de M larici­ populina, semblent donc être les voies les plus prometteuses dans la lutte contre ce champignon phytopathogène.

Comme nous l'avons vu, les applications des marqueurs microsatellites dans l'étude de la diversité génétique des populations d'agents pathogènes sont nombreuses et variées. Ces marqueurs permettent surtout d'évaluer des paramètres dont l'étude directe serait très difficile, voire impossible. Même si les marqueurs développés dans cette étude ont permis des avancées significatives dans la connaissance de la structuration et de la diversité génétique de M larici-populina, certaines limites ont pu être observées, notamment lors de la réalisation de tests d'assignation. Il est probable que le nombre de nos marqueurs soit insuffisant pour avoir la puissance nécessaire à ce type de tests. On peut également constater que le polymorphisme des marqueurs microsatellites isolés chez M larici-populina, même s'il est élevé, est bien en deçà de ce que l'on peut observer dans d'autres groupes taxonomiques. Ce polymorphisme relativement limité n'est pas l'apanage de cet agent pathogène, il semble en effet que le polymorphisme des loci microsatellites soit également faible dans d'autres espèces de champignons phytopathogènes (Dutech et al., soumis; annexe 1). Afin de renforcer la puissance et la qualité des études réalisées, il serait donc souhaitable de développer des marqueurs microsatellites supplémentaires. L'identification de nouveaux marqueurs pourra être grandement facilitée par le séquençage complet du génome de M larici-populina qui est en cours. De plus, la connaissance combinée des génomes complets d'un agent pathogène et de son hôte apportera très certainement de grandes connaissances sur la génétique de l'interaction entre le peuplier et M larici-populina, et fera très certainement ainsi de ce pathosystème un modèle.

126 Bibliographie Références bibliographiques

Agrios GN (1988) Plant Pathology, 803 pp. Academie Press, London.

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman Dl (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research, 25, 3389­ 3402.

Balloux F, Lehmann L, de Meeûs T (2003) The population genetics of clonaI and partially clonaI diploids. Genetics, 164,1635-1644.

Barnéoud C, Bonduelle P, Dubois JM (1982) Manuel de populiculture. Association Forêt-Cellulose, Paris, France.

Barrès B, Dutech C, Andrieux A, Caron H, Pinon J, Frey P (2006) Isolation and characterization of 15 microsatellite loci in the poplar rust fungus, Melampsora larici-populina, and cross-amplification in related species. Molecular Ecology Notes, 6, 60-64.

Barrett J (1985) The gene-for-gene hypothesis: parable or paradigm. In: Ecology and Genetics ofHost-parasite interactions (eds. Rollinson D, Anderson RM), pp. 215-225. Academie Press, London, UK.

Bart-Delabesse E, Humbert JF, Delabesse E, Bretagne S (1998) Microsatellite markers for typing Aspergillus fumigatus isolates. Journal ofClinical Microbiology 36,2413-2418.

Bastien C, Jorge V, Dowkiw A, Villar M, Faivre-Rampant P (2004) Amélioration génétique des peupliers. Biofutur, 247, 33-37.

Becker M, Picard JF, Timbal J (1982) Larousse des arbres, des arbustes et des arbrisseaux de l'Europe occidentale. Larousse, Paris, France.

Boileau MG, Hebert PDN, Schwartz SS (1992) Non-equilibrium gene frequency divergence: persistent founder effects in natural populations. Journal ofEvolutionary Biology, 5,25-39.

Bourassa M, Bernier L, Hamelin RC (2005) Direct genotyping of the poplar leaf rust fungus, Melampsora medusae f. sp. deltoidae, using codominant PCR-SSCP markers. Forest Pathology, 35, 245-261.

Bourassa M, Bernier L, Milligan BG, Hamelin RC (1998) Sympatry between alternate hosts affects population structure ofpoplar leafrust. Finnish Forest Research Institute, Research Paper, 712,65-69.

Bowden J, Gregory PH, Johnson CG (1971) Possible wind transport of coffee leaf rust across the Atlantic Ocean. Nature, 229, 500.

Brasier CM (2001) Rapid evolution ofintroduced plant pathogens via interspecific hybridization. Bioscience, 51, 123-133.

Brown JKM, Hovm011er MS (2002) Aerial dispersal of pathogens on the global and continental scales and its impact on plant disease. Science, 297, 537-541.

Brygoo Y, Caffier V, Carlier J, Fabre JV, Fernandez D, Giraud T, Mourichon X, Neema C, Notteghem JL, Pope C, Tharreau D, Lebrun MH (1998) Reproduction and population structure in phytopathogenic fungi. In: Molecular variability offungalpathogens (eds. Bridge PD, Couteaudier Y, Clarkson JM), pp. 133-148. CAB International, Wallingford, UK.

Burdon JJ (1992) Host population subdivision and genetic structure of natural pathogen populations. Advances in Plant Pathology, 8, 81-94.

Burdon JJ, Luig NH, Marshall DR (1983) Isozyme uniformity and virulence variation in Puccinia graminis f. sp. tritici and P. recondita f. sp. tritici in Australia. Australian Journal ofBiologicalSciences, 36, 403-410.

127 Références bibliographiques

Burdon JJ, Roberts JK (1995) The population genetic structure ofthe rust fungus Melampsora lini as revealed by pathogenicity, isozyme and RFLP markers. Plant Pathology, 44, 270-278.

Carlier J, Lebrun MH, Zapater MF, Dubois C, Mourichon X (1996) Genetic structure ofthe global population of banana leafstreak fungus, Mycosphaerellafljiensis. Molecular Ecology, 5, 499-510.

Chen RS, Boeger JM, McDonald BA (1994) Genetic stability in a population of a plant pathogenic fungus over time. Molecular Ecology, 3, 209-218.

Chesser RK, Baker RJ (1996) Effective sizes and dynamics of uniparentally and biparentally inherited genes. Genetics, 144, 1225-1235.

Chiba S, Khoda H (1985) Influence of leaf rust infection on photosynthetic activity in Populus maximowiczii A. HENRI. Oji Institutefor Forest Tree Improvement, 237, 151-153.

Cornuet JM, Luikart G (1996) Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics, 144,2001-2014.

Crowhurst RN, Hawthorne BT, Rikkerink EHA, Templeton MD (1991) Differentiation ofFusarium solani f. sp. cucurbitae races 1 and 2 by random amplification of polymorphie DNA. Current Genetics, 20, 391­ 396.

Dinoor A (1974) Role of wild and cultivated plants in the epidemiology of plant diseases in Israel. Annual Review ofPhytopathology, 12, 413-436.

Dowkiw A, Bastien C (2004) Characterization oftwo major genetic factors controlling quantitative resistance to Melampsora larici-populina leaf rust in hybrid poplars: strain specificity, field expression, combined effects, and relationship with a defeated qualitative resistance gene. Phytopathology, 94, 1358-1367.

Dowkiw A, Husson C, Frey P, Pinon J, Bastien C (2003) Partial resistance to Melampsora larici-populina leaf rust in hybrid poplars: genetic variability in inoculated excised leaf disk bioassay and relationship with complete resistance. Phytopathology, 93, 421-427.

Dutech C, Amsellem L, Billotte N, Jarne P (2000) Characterization of (GA)n microsatellite loci using an emichment protocol in the neotropical tree species Vouacapoua americana. Molecular Ecology, 9, 1433-1435.

Dutech C, Enjalbert J, Fournier E, Delmotte F, Barrès B, Carlier J, Tharreau D, Giraud T, Challenges in microsatellite isolation in fungi (soumis).

Eckenwalder JE (1996) Systemics and evolution of Populus. In: Biology ofPopulus and ifs implications for management and conservation (eds Stettler R, Bradshaw H, Hinckley T), pp. 7-30. NRC Research Press, Ottawa, Canada.

Edwards KJ, Barker JHA, Daly A, Jones C, Karp A (1996) Microsatellite libraries emiched for several microsatellite sequences in plants. Biotechniques, 20, 758-760.

Eisen J (1999) Mechanistic explanations for variation in microsatellite stability within and between species. In: Microsatellites: evolution and applications (eds. Goldstein D, SchlOtterer C), pp. 34--48. Oxford University Press, Oxford, England.

Ellegren H (2004) Microsatellites: simple sequences with complex evolution. Nature Reviews Genetics, 5, 435­ 445.

Enjalbert J, Duan X, Giraud T, Vautrin D, de Vallavielle-Pope C, Solignac M (2002) Isolation of twelve microsatellite loci, using an emichment protocol, in the phytopathogenic fungus Puccinia striiformis fsp tritici. Molecular Ecology Notes, 2, 563-565.

128 Ré@rences bibliographiques

Epperson BK, Huang Z, Li TQ (1999) Measures ofspatial structure in samples ofgenotypes for multiallelic loci. Genetica, 73, 251-261.

Estoup A, Solignac M, Harry M, Comuet JM (1993) Characterization of(GT)n and (CT)n microsatellites in two insect species: Apis mellifera and Bombus terrestris. Nucleic Acids Research, 21, 1427-1431.

Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: An integrated software package for population genetics data analysis. Evolutionary Bioinformatics Online, 1,47-50.

Feau N, Hamelin RC, Vandecasteele C, Stanosz GR, Bernier L (2005) Genetic structure of Mycosphaerella populorum (anamorph Septoria musiva) populations in North-Central and Northeastern North America. Phytopathology, 95, 608-616.

Felsenstein J (1989) PHYLIP - Phylogeny Inference Package (Version 3.2). Cladistics 5, 164-166.

Fitt BDL, McCartney HA, Walklate PJ (1989) The role of rain in dispersal of pathogen inoculum. Annual Review ofPhytopathology, 27, 241-270.

FIor HH (1942) Inheritance ofpathogenicity in Melampsora lini. Phytopathology, 32, 653-669.

FIor HH (1956) The complementary genic systems in flax and flax rust. Advances in Genetics, 8,29-54.

Foulon (1999) Etude de la variabilité moléculaire au sein des groupes de races E4 et E5 de Melampsora larici­ populina Kleb., agent de la rouille des peupliers: mise au point de la technique RAPD. Rapport de DAA, Institut National Agronomique Paris-Grignon, France, 39 p.

Fresa R (1936) Argentine Republic: Melampsora larici-populina in the Delta ofParana.International Bulletin of Plant Protection, 10, 145-146.

Frey P, Gérard P, Feau N, Husson C, Pinon J (2005) Variability and population biology ofMelampsora rusts on poplars. In: Rust diseases of Willow and Poplar (eds. Pei MH, McCracken AR), pp. 63-72. CAB International, Wallingford, UK.

Frey P, Pinon J (2004) La rouille du peuplier: un pathosystèmc modèle. Biofutur, 247, 28-32.

Fry WE, Goodwin SB, Matuszak JM, Spielman LJ, Milgroom MG, Drenth A (1992) Population genetics and intercontinental migrations ofPhytophthora irifestans. Annual Review ofPhytopathology, 30, 107-115.

Gastine F, Berthelot A, Bouvet A (2003) La protection phytosanitaire du cultivar 'Beaupré' est-elle efficace? Fiche Iriformations-Forêt Afoeel n0667, 6p.

Geagea L, Huber L, Sache 1 (1999) Dry-dispersal and rain-splash ofbrown (Puccinia recondita f.sp. tritici) and yellow (P. striiformis) rust spores from infected wheat leaves exposed to simulated raindrops. Plant Pathology, 48,472-482.

Geffroy V, Sicard D, de Oliveira JCF, Sevignac M, Cohen S, Gepts P, Neema C, Langin T, Dron M (1999) Identification of an ancestral resistance gene cluster involved in the coevolution process between Phaseolus vulgaris and its fungal pathogen Colletotrichum lindemuthianum. Molecular Plant-Microbe Interactions, 12, 774-784.

Gérard P, Husson C, Pinon J, Frey P (2006) Comparison of genetic and virulence diversity of Melampsora larici-populina populations on wild and cultivated poplar and influence of the alternate host. Phytopathology,96,1027-1036.

Gibson lAS, Waller lM (1972) Notes from ODA plant pathology liaison officers. Pest Articles and News Summaries, 18, 336-337.

129 Références bibliographiques

Giorcelli A, Vietto L (1998) Poplar and Marssonina: 35 years of cohabitation. Sherwood-Foreste ed Alberi Oggi, 4, 43-49.

Gobbin D, Jermini M, Loskill B, Pertot 1, Raynal M, Gessler C (2005) Importance of secondary inoculum of Plasmopara viticola to epidemies ofgrapevine downy mildew. Plant Pathology, 54, 522-534.

Goodwin PH, Annis SL (1991) Rapid identification of genetie variation and pathotype of Leptosphaeria maculans by Random Amplified Polymorphie DNA assay. Applied and Environmental Microbiology, 57,2482-2486.

Goudet J (1995) FSTAT (version 1.2): A computer program to calculate F-statistics. Journal ofHeredity, 86, 485­ 486.

Goudet J (2004) HIERFSTAT, a package for R to compute and test hierarchical F-statistics. Molecular Ecology Notes,5,184-186.

Goudet J, Raymond M, de Meeûs T, Rousset F (1996) Testing differentiation in diploid populations. Genetics, 144,1933-1940.

Groth JV, Roelfs AP (1982) Effect of sexual and asexual reprocuction on race abundance in cereal rust fungus populations. Phytopathology, 72, 1503-1507.

Grondin J, Bourassa M, Hamelin RC (2005) First report of the aecial state of Melampsora larici-populina on Larix spp. in North America. Plant Disease, 89, 1242.

Guérin F, Franck P., Loiseau A, Devaux M, Le Cam B (2004) Isolation of21 new polymorphie microsatellite loci in the phytopathogenic fungus Venturia inaequalis. Molecular Ecology Notes, 4,268-270.

Guérin F, Le Cam B (2004) Breakdown of the scab resistance gene Vf in apple leads to a founder effect in populations ofthe fungal pathogen Venturia inaequalis. Phytopathology, 94, 364-369.

Halkett F, Plantegenest M, Prunier-Leterme N, Mieuzet L, Delmotte F, Simon JC (2005) Admixed sexual and facultatively asexual aphid lineages at mating sites. Molecular Ecology, 14,325-336.

Hancock JM (1995) The contribution of slippage-like processes to genome evolution. Journal of Molecular Evolution, 41, 1038-1047.

Hartl DL & Clark AG (1997) Principles ofPopulation Genetics. 3rd edn. Sinauer Associates, Sunderland, USA.

Hermansen JE, Torp U, Prahm LP (1978) Studies of transport of live spores of cereal mildew and rust fungus across the North Sea. Grana, 17,41-46.

Hovmoller MS, Justesen AF, Brown JKM (2002) Clonality and long distance migration ofPuccinia striiformis f. sp. tritici in north-west Europe. Plant Pathology, 51, 24-32.

Innes L, Marchand M, Frey P, Bourassa M, Hamelin RC (2004) First report of Melampsora larici-populina on Populus spp. in Eastern North America. Plant Disease, 88, 85.

Jarne P, Lagoda PJL (1996) Microsatellites, from molecules to populations and back. Trends in Ecology and Evolution, 11, 424-429.

Joosten MHAJ, Cozijnsen AI, De Wit PJGM (1994) Host resistance to a fungal tomato pathogen lost by a single base-pair change in an avirulence gene. Nature, 367, 384-387.

130 Références bibliographiques

Jorge V, Dowkiw A, Faivre-Rampant P, Bastien C (2005) Genetic architecture of qualitative and quantitative Melampsora larici-populina leaf rust resistance in hybrid poplar: genetic mapping and QTL detection. New Phytologist, 167, 113-127.

Karaog1u H, Lee CMY, Meyer W (2005) Survey of simple sequence repeats in completed funga1 genomes. Molecular Biology andEvolution 22,1159-1159.

Kashi Y, Soller M (1999) Functiona1 roles of microsatellites and minisatellites. In: Microsatellite evolution and application (eds. Goldstein DD, Schlotterer C), pp. 10-23. Oxford University Press, Oxford, UK.

Kern FD, Thurston HW (1954) Additional species ofUredinales from Colombia. Mycologia, 46, 354-357.

Kolmer JA (1991a) Evolution of distinct populations of Puccinia recondita f. sp. tritici in Canada. Phytopathology, 81, 316-322.

Ko1mer JA (1991b) Phenotypic diversity in two populations ofPuccinia recondita f. sp. tritici in Canada during 1931-1987. Phytopathology, 81, 311-315.

Kumar S, Tamura K, Nei M (2004) MEGA3: Integrated software for Mo1ecu1ar Evolutionary Genetics Analysis and sequence alignment. Briefings in Bioinformatics, 5, 150-163.

Kusha1appa AC, Eskes AB (1989) Advances in coffee rust research. Annual Review ofPhytopathology, 27, 503­ 531.

Laurans F (1997) DéfInition histologique et approche moléculaire des interactions peupliers-rouilles. Thèse de doctorat, Université d'Orléans, France, 138 p.

Laurans F, Pilate G (1999) Histologica1 aspects of a hypersensitive response in poplar to Melampsora larici­ populina. Phytopathology, 89, 233-238.

Legionnet A (1996) Diversité et fonctionnement génétique des populations naturelles de Populus nigra 1., espèce pionnière des ripisy1ves Européennes. Thèse de doctorat, Université de Montpellier II, France, 120 p.

Legionnet A, Lefèvre F (1996) Genetic variation in the riparian pioneer tree species Populus nigra 1. 1. Study of population structure based on isozymes. Heredity, 77, 629-637.

Leonard KJ, Roelfs AP, Long DL (1992) Diversity of virulence within and among populations of Puccinia recondita f. sp. tritici in different areas ofthe United States. Plant Disease, 76, 500-504.

Levinson G, Gutman GA (1987) Slipped-strand mispairing: a major mechanism for DNA sequence evolution. Molecular Biology andEvolution 4, 203-221.

Li YC, Korol AB, Fahima T, Beiles A, Nevo E (2002) Microsatellites: genomic distribution, putative functions and mutational mechanisms: a review. Molecular Ecology, 11,2453-2465.

Li CV, Li JB, Zhou XG, Dong AR, Xu MH (2004) Frequency and distribution of microsatellites in the whole genome ofrice blast fungus, Magnaporthe grisea. Chinese Journal ofRice Science, 18,269-273.

Lim S, Notley-McRobb L, Lim M, Carter DA (2004) A comparison of the nature and abundance of microsatellites in 14 fungal genomes. Fungal Genetics andBiology, 41, 1025-1036.

Linde CC, Zhan J, McDonald BA (2002) Population structure of Mycosphaerel!a graminicola: From lesions to continents. Phytopathology, 92, 946-955.

Lloyd HL (1971) New records ofRhodesian fungi. 1. Melampsora larici-populina Klebahn. Kirkia, 8, 1-2.

131 Références bibliographiques

Long DL, Roelfs AP, Leonard KJ (1993) Virulence and diversity ofPuccinia recondita f. sp. tritici in the United States in 1991. Plant Disease, 77, 786-791.

Lonsdale D, Tabbush P (2002) Poplar rust and its recent impact in Great Britain. Information Note 7 (revised). Forestry Commission. Edimburgh, UK.

Luikart G, England PR (1999) Statistical analysis ofmicrosatellite DNA data. Trends in Ecology and Evolution, 14,253-256.

Luikart G, England PR, Tallmon D, Jordan S, Taberlet P (2003) The power and promise of population genomics: from genotyping to genome typing. Nature Reviews Genetics 4,981-994.

Mantel N (1967) The detection of disease clustering and a generalized regression approach. Cancer Research, 27, 209-220.

Martin F, Kohler A (2004) Génomique structurale et fonctionnelle du peuplier. Biofutur, 247, 38-42.

Maynard Smith J, Smith NH, O'Rourke M, Spratt BG (1993) How clonai are bacteria? Proceedings of the National Academy ofSciences ofthe USA, 90,4384-4388.

McDermott JM, McDonald BA (1993) Gene flow in plant pathosystems. Annual Review ofPhytopathology, 31, 353-373.

McDonald BA (1997) The population genetics offungi: tools and techniques. Phytopathology, 87, 448-453.

McDonaldBA, Linde CC (2002) The population genetics of plant pathogens and breeding strategies for durable resistance. Euphytica, 124, 163-180. l'v1cDonald BA, Martinez JP (1990) DNA restriction fragment length polymorphisms in Mycosphaerella graminicola (anamorph Septoria tritici) isolates collected from a single wheat field. Phytopathology, 80,1368-1373.

Metzgar D, Bytof J, Wills C (2000) Selection against frameshift mutations limits microsatellite expansion in coding DNA. Genome Research, 10, n-8û.

Miot S, Frey P, Pinon J (1999) Varietal mixture of poplar clones: Effects of infection by Melampsora larici­ populina and on plant growth. European Journal ofForest Pathology, 29, 411-423.

Mullis KB, Faloona FA (1987) Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods in Enzymology, 155,335-350.

Nagarajan S, Singh DV (1990) Long-distance dispersal of rusts pathogens. Annual Review ofPhytopathology, 28,139-153.

Nei M (1972) Genetic distance between populations. The American Naturalist, 106,283-292.

Neu C, Kaemmer D, Kahl G, Fischer D, Weising K (1999) Polymorphie microsatellite markers for the banana pathogen Mycosphaerellafijiensis. Molecular Ecology 8,523-525.

Newcombe G (1998) A review of exapted resistance to disease of Populus. European Journal of Forest Pathology, 28, 209-216.

Newcombe G, Chastagner GA (1993) First report of the Eurasian poplar leaf rust fungus, Melampsora larici­ populina, in North-America. Plant Disease, 77, 532-535.

132 Références bibliographiques

Niemi L, Wennstrom A, Hjaltén J, Waldmann P, Ericson L (2006) Spatial variation in resistance and virulence in the host-pathogen system Salix triandra-Melampsora amygdalinae. Journal ofEcology, 94, 915-921.

Olafsson E, Bjôrnsson H (1997) Fidrildi a Islandi 1995. Fjolrit NattUrufraedistofnunar No 32 [Lepidoptera in Iceland 1995. Report ofthe Icelandic Institute ofNatural History No 32], 138 pp.

Page RDM (1996) TREEVIEW: an application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences, 12, 357-358.

Parker MA (1985) Local population differentiation for compatibility in an annual legume and its host-specifie fungal pathogen. Evolution, 39, 713-723.

Peakall R, Ruibal M, Lindenmayer DB (2003) Spatial autocorrelation analysis offers new insights into gene flow in the Australian bush rat, Rattusfuscipes. Evolution, 57, 1182-1195.

Peakall R, Smouse PE (2005) GenAlEx 6: genetic analysis in Excel. Population genetic software for teaching and research. Molecu/ar Ecology Notes, 6, 288-295.

Pei MH, Ruiz C, Bayon C, Hunter T, Lonsdale D (2005) Pathogenic variation in poplar rust Melampsora larici­ populina from England. European Journal ofPlant Pathology, 111, 147-155.

Pei MH, Whelan MJ, Halford NG, Royle Dl (1997) Distinction between stem- and leaf-infecting forms of Melampsora rust on Salix viminalis using RAPD markers. Mycological research, 101, 7-10.

Petit RJ, Deguilloux MF, Chat J, Grivet D, Garnier-Géré P, Vendramin GG (2005) Standardizing for microsatellite length in comparisons ofgenetic diversity. Molecular Ecology 14,885-890.

Pinon J, Frey P (1997) Structure of Melampsora larici-populina populations on wild and cultivated poplar. European Journal ofPlant Pathology, 103,159-173.

Pinon J, Frey P (2005) Interactions between poplar clones and Melampsora populations and their implications for breeding for durable resistance. In: Rust diseases ofWillow and Poplar (eds. Pei MH, McCracken AR), pp. 139-154. CAB International, Wallingford, UK.

Pinon J, Frey P, Husson C (2006) Wettability of poplar leaves influences dew formation and infection by Melampsora larici-populina. Plant Disease, 90, 177-184.

Pinon J, Frey P, Husson C, Schipfer A (1998) Poplar rust (Melampsora larici-populina): the development ofE4 pathotypes in France since 1994. Finnish Forest Research Institute, Research Papers, 712, 57-64.

Pinon J, Newcombe G, Chastagner GA (1994) Identification of races of Melampsora larici-populina, the Eurasian rust fungus, on Populus species in California and Washington. Plant Disease, 78, 101.

Piry S, Luikart G, Cornuet JM (1999) BOTTLENECK: a computer program for detecting recent reductions in the effective population size using allele frequency data. Journal ofHeredity, 90, 502-503.

Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics, 155, 945-959.

Purdy LH, Krupa SV, Dean JL (1985) Introduction ofsugarcane rust into the Americas and its spread to Florida. Plant Disease, 69, 689-693.

R Development Core Team (2004) R: a language and environment for statistical computing. R foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org.

133 Références bibliographiques

Ranum LP, Day JW (2002) Dominantly inherited, non-coding microsatellite expansion disorders. Current Opinion in Genetics & Development, 12, 266-271.

Raymond M, Rousset F (1995) GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. Journal ofHeredity, 86, 248-249.

Rice WR (1989) Analyzing tables ofstatistical tests. Evolution, 43, 223-225.

Rivas GG, Zapater MF, Abadie C, Carlier J (2004) Founder effects and stochastic dispersal at the continental scale ofthe fungal pathogen ofbananas Mycosphaerellafzjiensis. Molecular Ecology, 13,471-482.

Robert C, Bancal MO, Lannou C (2002) Wheat leaf rust uredospore production and carbon and nitrogen export in relation to lesion size and density. Phytopathology, 92, 762-768.

Roelfs AP, Groth JV (1980) A comparison of virulence genotypes in wheat stem rust populations reproducing sexually and asexuaIly. Phytopathology, 70, 855-862.

Rousset F (1997) Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Genetics, 145, 1219-1228.

Rozen S, Skaletsky H (2000) Primer 3 on the WWW for general users and for biologist programmers. In: Bioiriformatics Methods andProtocols: Methods in Molecular Biology (eds. Krawetz S, Misener S), pp. 365-386. Humana Press, Totowa, New Jersey.

Sales C (1995) Le peuplier dans la filière bois. Compte-Rendus de l'Académie d'Agriculture de France, 81, 15­ 22.

Samils B, Stepien V, Lagercrantz D, Lascoux M, Gullberg U (2001) Genetic diversity in relation to sexual and asexual reproduction in populations ofMelampsora larici-epitea. European Journal ofPlant Pathology, 107,871-881.

Selkoe KA, Toonen RJ (2006) Microsatellites for ecologists: a practical guide to using and evaluating microsatellite markers. Ecology Letters, 9, 615-629.

Slatkin M (1977) Gene flow and genetic drift in a species subject to frequent local extinctions. Theoretical Population Biology, 12, 253-262.

Smouse PE, Peakall R (1999) Spatial autocorrelation analysis of individual multiallele and multilocus genetic structure. Heredity, 82, 561-573.

Spegazzini C (1918) Cuarta contribuci6n a la Micologia chilena. Revista Chilena Historia Natural, 22, 95-104.

Spiers AG, Hopcroft DH (1988) Penetration and infection of poplar leaves by urediniospores of Melampsora larici-populina and Melampsora medusae. New Zealand Journal ofBotany, 26, 101-111.

Steimel L, Chen W, Harrington TC (2005) Development and characterization of microsatellite markers for the poplar rust fungi Melampsora medusae and Melampsora larici-populina. Molecular Ecology Notes, 5, 484-486.

Stenberg P, Lundmark M, Saura A (2003) MLGsim: a program for detecting clones using a simulation approach. Molecular Ecology Notes, 3, 329-331.

Stephens M, Smith NJ, Donnelly P (2001) A new statistical method for haplotype reconstruction from population data. The American Journal ofHuman Genetics, 68, 978-989.

134 Ré{erences bibliowaphiques

Stettler RF, Zsuffa L, Wu R (1996) The role ofhybridization in the genetic manipulation ofPopulus. In: Biology of Populus and its implications for management and conservation (eds Stettler R, Bradshaw H, Hinckley T), pp. 87-112. NRC Research Press, Ottawa, Canada.

Stoddart JA, Taylor JF (1988) Genotypic diversity: estimation and prediction in samples. Genetics, 118, 705­ 711.

Sujkowski LS, Goodwin SB, Dyer AT, Fry WE (1994) Increased genotypic diversity via migration and possible occurrence ofsexual reproduction ofPhytophthora infestans in Poland. Phytopathology, 84, 201-207.

Sunnucks P (2000) Efficient genetic markers for population biology. Trends in Ecology and Evolution, 15, 199­ 203.

Taylor JW, Jacobson DJ, Fisher MC (1999) The evolution of asexual fungi: reproduction, speciation and classification. Annual Review ofPhytopathology, 37, 197-246.

Thrall PH, BUïdon JJ (2002) Evolution of gene-for-gene systems in metapopulations: the effect of spatial scale ofhost and pathogen dispersal. Plant Pathology, 51, 169-184.

Thrall PH, Burdon JJ (2003) Evolution of virulence in a plant host-pathogen metapopulation. Science, 299, 1735-1737.

Tooley PW, Fry WE, Villareal MJ (1985) Isoenzyme characterization of sexual and asexual Phytophthora infestans population. Journal ofHeredity, 76,431-435.

T6th G, Gaspari Z, Jurka J (2000) Microsatellites in different eukaryotic genomes: survey and analysis. Genome Research 10, 967 - 981.

Turner ME, Stephens JC, Anderson WW (1982) Homozygosity and patch structure in plant populations as a result ofnearest-neighbor pollination. Proceedings ofthe National Academy ofSciences ofthe USA, 79, 203-207.

Tuskan GA, DiFazio SP, Hellsten U, Jansson S, Rombauts S, Putnam N, Sterck L, Bohlmann J, Schein J, Bhalerao RR, Bhalerao RP, Blaudez D, Boerjan W, Brun A, Brunner A, Busov V, Campbell M, Carlson J, Chalot M, Chapman J, Chen G, Cooper D, Coutinho PM, Couturier J, Covert SF, Cunningham R, Davis J, Degroeve S, dePamphilis C, Detter J, Dirks B, Dubchak l, Duplessis S, Ehlting J, Ellis B, Gendler K, Goodstein D, Gribskov M, Grigoriev 1, Groover A, Gunter L, Hamberger B, Heinze B, Helariutta Y, Henrissat B, Holligan D, Islam-Faridi N, Jones-Rhoades M, Jorgensen R, Joshi C, Kangasjarvi J, Karlsson J, Kelleher C, Kirkpatrick R, Kirst M, Kohler A, Kalluri U, Larimer F, Leebens-Mack J, Leplé JC, Dejardin A, Pilate G, Locascio P, Lucas S, Martin F, Montanini B, Napoli C, Nelson DR, Nelson CD, Nieminen KM, Nilsson 0, Peter G, Philippe R, Poliakov A, Ralph S, Richardson P, Rinaldi C, Ritland K, Rouzé P, Ryaboy D, Salamov A, Schrader J, Segerman B, Sterky F, Souza C, Tsai C, Unneberg P, Wall K, Wessler S, Yang G, Yin T, Douglas C, Sandberg G, Van de Peer Y, Rokhsar D. 2006. The genome of black cottonwood, Populus trichocarpa (Torr. & Gray). Science, 313, 1596-1604.

Valière N (2002) GIMLET: a computer program for analysing genetic individual identification data. Molecular Ecology Notes, 2, 377-379.

Vanden Broeck A, Storme V, Cottrell JE, Boerjan W, Van Bockstaele E, Quataert P, Van Slycken J (2004) Gene f10w between cultivated poplars and native black poplar (Populus nigra L.): A case study along the river Meuse on the Dutch-Belgian border. Forest Ecology and Management, 197,307-310.

Vanden Broeck A ,Villar M,Van Bockstaele E ,Van Slycken J (2005) Natural hybridization between cultivated poplars and their wild relatives: evidence and consequences for native poplar populations. Annals of Forest Science, 62, 601-613.

135 Références bibliographiques

Villar M, Le Bouler H, Forestier 0 (2004) Le peuplier noir: un patrimoine naturel à preserver. Biofutur, 247, 24­ 27.

Villar M, Lefèvre F, Augustin S, Bonduelle P, Delplanque A, Duval H, Faivre-Rampant P, Goué MC, Legionnet A, Ménard M, Nesme X, Pilate G, Pinon J, Valadon A, Teissier du Cros E (1995) Perspectives d'amélioration génétique des peupliers en France. Compte-Rendus de l'Académie d'Agriculture de France, 81,137-151.

Walker J, Hartigan D, Bertus AL (1974) Poplar Rusts in Australia with comments on potential conifer rusts. European Journal ofForest Pathology, 4, 100-118.

Wang J, van der Kamp BJ (1992) Resistance, tolerance, and yield of western black cottonwood infected by Melampsora rust. Canadian Journal ofForest Research, 22, 183-192.

Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358-1370.

Wilkinson AG, Spiers AG (1976) Introduction ofthe poplar rusts Melampsora larici-populina and M medusae to New Zealand and their subsequent distribution. New Zealand Journal ofScience, 19, 195-198.

Wisler GC, Norris RF (2005) Interactions between weeds and cultivated plants as related to management of plant pathogens. Weed Science, 53, 914-917.

Xu X, Peng M, Fang Z (2000) The direction ofmicrosatellite mutations is dependent upon allele length. Nature Genetics, 24, 396-399.

Yang RC (1998) Estimating hierarchical F-statistics. Evolution, 52, 950-956.

Zane L, Bargelloni L, Patarnello T (2002) Strategies for microsatellite isolation: a review. Molecular Ecology, 11, 1-16.

136 Annexe 1 Développement de microsatellites chez les champignons

Challenges of microsateUite isolation in fungi

1 2 3 CYRIL DUTECH , JEROME ENJALBERT , ELISABETH FOURNIER , FRANÇOIS 4 5 6 6 DELMOTTE , BENOIT BARRES , JEAN CARLIER , DIDIER THARREAU , and TATIANA GIRAUD 7

1 Biodiversité, Gènes et Communautés, INRA - Université Bordeaux 1, Centre de Bordeaux, 33612 Cestas, France

2 Laboratoire de Pathologie Végétale, INRA, 78850 Thiverval Grignon, France

3 Phytopathologie et Méthologie de la Détection Versailles, INRA, Route de Saint Cyr, 78026 Versailles, France

4 Santé Végétale, INRA -ENITA Bordeaux, Centre de Bordeaux, BP 81,33883 Villenave d'Ornon, France

5 Interactions Plantes-Microorganismes, INRA-Nancy, 54280 Champenoux, France

6 Biologie et Génétique des Interactions Plantes-Parasites, CIRAD-INRA-AGROM, TA 41/K, 34398 Montpellier, France

7 Ecologie, Systématique et Evolution CNRS- Université Paris-Sud, Bâtiment 360, Université Paris-Sud, 91405 Orsay, France

Keywords: SSR, plant pathogen, polymorphie markers, development, isolation, features affeeting polymorphism

Running title: Fungal microsatellites

Corresponding author: Cyril Duteeh. INRA-Bordeaux, UMR BIOGECO, Equipe de Pathologie Forestière, Domaine de la Grande Ferrade, BP81, 33883 Villenave d'Ornon Cedex, France [email protected] Tel: +33 5 57 122624. Fax: +33 557 122621

seo UHP 1\11'1\1cY l Biblir~~a~ue des Sciences R~e d:è;~:~in Botanique - CS 2014g 137 54601 VILLER$ LES NANCY omEX Annexe 1

Abstract

Although they represent powerful genetic markers in many fields of biology, microsatellites have been isolated in few fungal species. The aim of this study was to assess whether obtaining microsatellite markers with an acceptable level ofpolymorphism is generally harder in fungi than in other organisms. We therefore surveyed the number, nature and polymorphism level of published microsatellite markers in fungi from the literature and from our own data on seventeen fungal microsatellite-enriched libraries. Compared to microsatellites of five other phylogroups (angiosperms, insects, fishes, birds and mammals), fungal microsatellites indeed appeared both harder to isolate ,md to exhibit lower polymorphism. This low polymorphism appeared to be due, at least in part, to genomic specificities, such as scarcity and shortness of fungal microsatellite loci. A correlation was indeed observed between mean repeat number and mean allele number in the published microsatellite loci. The cross-species transferability of fungal microsatellites also appeared lower than in other phylogroups. However, microsatellites have been useful in sorne fungal species and the huge advantages of these markers render them nevertheless worth trying to develop in fungi. This study provides sorne guidelines for microsatellite isolation. We advocate that, if the focal species is known or suspected to have an overalllow diversity and that difficulties are met in the first enriched library, alternatives should be rapidly sought that should prove more profitable.

138 Développement de microsatellites chez les champignons

Introduction

Microsatellite loci, short tandemly repeated motifs of 1-6 bases, also known as simple sequence repeats (SSR), are widely used as genetic markers because oftheir ubiquity, ease to score, co-dominance, reproducibility, assumed neutrality and high level of polymorphism {Jarne, 1996 #1}. They proved to be invaluable in many fields of biology, from genome mapping to forensics, paternity testing and population genetics {Jarne, 1996 #1; Luikart, 2003 #12}. Their interest goes even beyond their high polymorphism: when one can assume a model for their evolution, taking into account the number of repeats allows inferring kin relationships among alleles and thus developing powerful tools for inferring evolutionary and demographic parameters {Cornuet, 1999 #14; Luikart, 2003 #12; Michalakis, 1996 #15}. The major drawback of microsatellite loci is that they often need to be isolated de novo in each species, which can be time-consuming and expensive. Cross-amplification, i.e. amplification ofloci in another species than the one in which they were cloned, is indeed generally possible only among species of the same genera, and even in this case the percentage of cross­ amplification is low {Rossetto, 2001 #98}. Furthermore, cross-amplification often generates numerous null alleles which can bias genetic analyses {Hardy, 2003 #106}. In species for which no microsatellite markers for related species are available or cross-amplifiable, recently developed techniques, especially those involving enrichment of genomic DNA in microsatellites {Zane, 2002 #2}, have rendered the step of microsatellite isolation less laborious and more likely to succeed. However, the task of developing a working primer set from an enriched library can also represent a significant workload {Squirrell, 2003 #84}.

Microsatellites have been isolated across a wide range of taxonomie groups, but surprisingly little in fungi {Zane, 2002 #2}; Fig. 1). The low number ofpopulation geneticists interested in fungi compared to other organisms certainly explains this rarity, in addition to a preference for anonymous markers such as random amplified polymorphie DNA markers (RAPD), amplified fragment length polymorphisms (AFLP) and inter-simple sequence repeats (ISSR). These markers have only two alleles per locus, but they are easy to develop in large numbers without the fastidious step of building a genomic library and they generally yield enough polymorphism to differentiate individuals within populations. However, for sorne fungal species, their lack of species specificity can represent a serious problem. For instance in fungal pathogens, DNA of the focal species can be difficult to isolate from those of the host and ofhyperparasites (e.g. mycoparasites; Kiss 1998), thus requiring an in vitro isolation step.

139 Annexe J

Other drawbacks of AFLP, RAPD and ISSR include dominance, preventing the detection of heterozygotes in diploid species. Furthermore, even for haploid species convenient for in vitro culture, the problem of anonymity remains, which can introduce serious bias in population genetic studies. Indeed, different alleles from a single locus cannot be easily recognized and markers with occasional non mendelian behaviour, such as transposable elements, are frequently amplified by these techniques. Other advantages of microsatellites include the possibility ofusing minute amount ofDNA, again allowing skipping a culture step, and easier detection ofamplification problems, such as nUll alleles.

In addition to the low number of population studies on fungi and the preference for anonymous markers, sorne peculiar genomic and biological traits of fungi may have limited the number of polymorphie microsatellite loci isolated from genomic libraries, thus contributing to the low number of published microsatellite developments. First, pathogens which are the most extensively studied species within fungi, have demographic and reproductive traits promoting a low genetic diversity. Crop or human pathogens have for instance often experienced recent bottlenecks, through geographical introduction {Rivas, 2004 #22; Engelbrecht, 2004 #23; Milgroom, 1992 #25} or host shifts {Paraskevis, 2003 #19; Tobler, 2003 #20; Mackenzie, 2001 #21}, which can drastically reduce intraspecific genetic diversity. Furthermore, sorne specifie life history traits of fungal pathogens, such as frequent asexual reproduction and recurrent bottlenecks in epidemic cycles, associated with low winter survival and/or selective sweeps following new virulences, are also likely to result in low level of genetic diversity {Goodwin, 1994 #24; Hovm011er, 2002 #26; Guérin, 2004 #27}. Second, fungal genomes may exhibit sorne peculiarities, such as a rarity of microsatellites, a low level of their polymorphism due to low numbers of repeats or a predominance of loci with a low mutation rate, for instance those with imperfect repeats. Several recent papers have examined the nature and abundance of microsatellites in published complete fungal genomes {Field, 1998 #3; Karaoglu, 2005 #4; Lim, 2004 #5}. Microsatellites indeed appeared less abundant in these fungal genomes than in other organisms {T6th, 2000 #11; Morgante, 2002 #18}, had different most abundant motifs {T6th, 2000 #11; Morgante, 2002 #18} and long loci were under-represented {Lim, 2004 #5; Karaoglu, 2005 #4}. {Lim, 2004 #5} reported for instance that ca. 90% of microsatellite loci in 14 fungal genomes had low numbers ofrepeats, i.e. below seven. Several studies have shown that the number ofrepeats is a good predictor of the level ofvariability in other organisms (e.g. {Goldstein, 1995 #70}. Ifthis correlation holds in fungi, most of their microsatellites are expected to exhibit a low polymorphism.

140 Développement de microsatellites chez les champignons

Furthermore, most of the microsatellite loci detected in the published fungal genomes (94%) were mononucleotide repeats {Lim, 2004 #5} that are seldom used in population genetics because of difficulties in scoring alleles separated by single base pairs. The shortness of microsatellite loci in fungi, their weak representation in the genomes, the low abundance of useful motifs, together with the small size offungal genomes (between 10 and 40 Mb), should limit the ability to find numerous polymorphic microsatellite markers by screening a genomic library.

However, it is not clear whether the genomic and biological specificities listed above really impair the development of microsatellite markers in fungi. First, the conclusions drawn from these genomic studies are limited by the low number ofcomplete fungal genomes available, a huge variability having been reported among closely related species in the number and nature of microsatellites {Ellegren, 2004 #13; Lim, 2004 #5; Karaoglu, 2005 #4}. A survey of microsatellite development studies in different fungal species would allow determining whether microsatellites are indeed generally difficult to isolate and are particularly short. Second, another important limitation of genome analyses is the lack of polymorphism assessment, which is the most valuable information for population geneticists. Estimations of demographic or genetic parameters are indeed more powerful with more polymorphic loci (e. g. {Paetkau, 2004 #99}. If for instance microsatellite loci with short repeats are reasonably polymorphic in fungi, their predominance in the genomes would not be a problem for the development of useful markers. Comparing the degree of polymorphism of microsatellites in fungi and in other organisms and assessing whether the correlation between the number of alleles and the number of repeats holds in fungi would therefore be crucial to determine whether attempts to develop microsatellites in this kingdom are worthwhile given the investment required.

The aim of this paper was therefore to assess the yield of enriched libraries in fungi and to compare the polymorphism of isolated fungal microsatellites to that of other organisms, to determine whether obtaining microsatellite markers with an acceptable level ofpolymorphism is generally harder in fungi than in other organisms. The specifie objectives ofthis paper were thus to (1) assess the yield of our own seventeen microsatellite-enriched libraries, through the different steps, to identify which ones limited the isolation of polymorphie loci; our data are free from publication bias, whereas failures to develop polymorphie markers are rarely published; (2) estimate the general yield ofpublished microsatellite development in fungi; (3)

141 Annexe 1

evaluate the possibility ofcross-transferability ofmicrosatellites among fungal species, which may represent an alternative to the fastidious development of a genomic library; (4) assess whether there is a correlation between length and allele number among fungal microsatellites; (S) compare the nature, in particular the size, of fungal microsatellites and their level of polymorphism to those of other groups of organisms. In this study, we considered fungal species sensus lato, i.e. including Oomycota because these organisms share intriguing similarities with true fungi in their morphology and life cycles, and many are also responsible for destructive plant diseases {Tyler, 2001 #8S}.

MateriaI and methods

Enriched libraries

The methods used to isolate our microsatellite loci were adapted from two protocols using oligoprobes for the enrichment of genomic libraries. The principle of both methods is the hybridization of restricted genomic DNA on microsatellite oligoprobes, followed by the washing of the non-hybridized genomic fragments. The first protocol, adapted from {Edwards, 1996 #88}, uses membranes on which microsatellite oligoprobes are fixed. The second method is very similar, but uses streptavidin-coated magnetic beads on which biotin­ labelled microsatellite oligoprobes are linked. Genomic fragments containing microsatellites hybridize with the oligoprobes, whose biotin links to the streptavidin ofthe magnetic beads. A magnet therefore allows retaining mainly the DNA fragments with microsatllite loci {Kijas, 1994 #94}. The first method, with a membrane, was used for the species Cryphonectria parasitica {Breuillin, 2006 #87}, Erysiphe alphitoides(unpublished data) and Melampsora larici-populina {Barrès, 200S #S8}. The bead method was used in addition for E. alphitoides and M larici-populina, and for the 14 other species: Erysiphe necator (unpublished data), Fusarium culmorum {Giraud, 2002 #7}, Fusarium poae (unpublished data), Magnaporthe grisea {Kaye, 2003 #57}, Microbotryum violaceum {Giraud, 2002 #10}, Microcyclus ulei {Le Guen, 2004 #33}, Mycosphaerella eumusae (unpublished data), Mycosphaerella fijiensis (unpublished), Mycosphaerella musicola (unpublished), Penicillium camembertii (unpublished data), Penicillium roqueforti (unpublished data), Plasmopara viticola {Delmotte, 2006 #86}, Puccinia triticina {Duan, 2003 #9} and Puccinia striiformis fsp tritici {Enjalbert, 2002 #8}. For sorne species, severallibraries had to be performed because of the poor yield ofthe first one(s).

142 Développement de microsatellites chez les champignons

Twelve out of 17 librairies (70%) were enriched for dinucleotide loci using (AC/TG)n and/or (AG/TC) n oligoprobes, with n= 10 or 15 (Table 1). The last five librairies were enriched using only (AC)IO. These two dinucleotide motifs were chosen for enrichment because they had generally been reported as the most frequent in complete fungal genomes (e.g. {Lim, 2004 #5}. Ten librairies were also enriched with trinucleotide motifs (mainly (AAG)IO; Table 1) to increase the number ofpolymorphic loci isolated.

For each of our enriched libraries, we recorded: (1) the percentage of positive clones, (2) the percentage of redundant sequences, i.e. of identical sequences, (3) the percentage of contaminant clones, i.e. with a significant blast towards a sequence from another species, (4) the percentage of unique sequences, excluding contaminants, with a microsateIlite locus (tandemly repeated motifs of 1-6 bases with at least 5 pure repeats, according to the most common definition; {Ashley, 1994 #89; Lim, 2004 #5}, (5) the percentage of unique sequences, excluding contaminants, with a microsatellite locus and suitable flanking regions, (6) the percentage of sequences yielding loci with a clear amplification, (7) the percentage of sequences yielding polymorphic loci at the intra-population level and (8) the percentage of sequences yielding polymorphic loci at the largest measured scale (from inter-population to inter-continental levels, or between populations from different host species). AIl the above percentages were estimated as ratios over the number of inserts correctly sequenced, except the percentage of positive clones, which was estimated over the total number of clones with inserts. When several libraries had been built for one species, the average yield was taken for each step. In addition, we recorded for each polymorphic locus: (l) The base composition of the microsatellite motif (2) its perfection (a locus was considered as imperfect if the tandem repeats were interrupted or if several different tandem repeats with more than five repeats each were amplified as a single locus) and (3) the number of tandem repeats (for imperfect loci, number of repeats of the longest perfect microsatellite). The number of repeats was recorded from the sequenced fragment obtained in the library. FinaIly, we estimated genetic diversity of microsateIlite loci as the number of alleles at the largest scale (spatiaIly or inter­ hosts). We also recorded the sample size used for assessing polymorphism.

Literature search and data extraction

143 Annexe J

To find studies reporting the development of microsatellite markers in fungi, we searehed the bibliographie data bases Web of Knowledge (http://isi4.newisiknowledge.com/) and Pubmed (http://www.nebi.nlm.nih.gov/entrez/query.fcgi) from January 1985 to June 2006 for aIl papers with "mierosat* and fung* and (isol* or clon* or eharaeteri*)" and "SSR and fung* and (isol* or clon* or eharaeteri*)" in the title, keyword or abstraet. We included aIl data from aIl papers to whieh we had aeeess, regardless ofthe method of microsatellite isolation, except that we kept a single study per speeies and only the studies having isolated at least two polymorphie dinucleotide loci. We kept a single study per speeies and only the studies having isolated at least two polymorphie dinucleotide loci in addition to the other loci (tri-,tetra-... nucleotides), in order to perform polymorphism comparisons only based on the dinucleotide loci (see below). For each locus, we recorded the following information when available (1) the length and base composition of the motif, (2) its perfection, (3) the number of repeats of the longest perfeet microsatellite (4) the sample size used for assessing polymorphism, (5) the number ofalleles. We also reeorded the number loci for whieh primers eould be designed, the percentage ofseorable loci and the pereentage polymorphie loei per species, when available.

In addition, eross-speeies transferability ofmicrosatellite markers in fungi was evaluated from published studies and our own data. We kept 20 studies for whieh one source speeies eould be cleady identified and data on eross-speeies transferability was available within a genus, data at lower or higher taxonomie levels being searee in fungi. In total, 24 source speeies, 88 target speeies and 302 primer pairs were tested aeross these studies. For eaeh target speeies, a primer pair was eonsidered as transferable when a PCR produet ofexpeeted size was obtained in at least one individual. We eomputed the transferability as the mean percentage of loei that were transferable to other species.

To compare the yield and polymorphism of microsatellite development in fungi to those of other organisms, we searched in the issues of Molecular Ecology Notes from March 2001 to June 2005 the studies reporting isolation of microsatellites in angiosperms, insects, fishes (restricted here to Actinopterygii), birds and mammals. These different phylogenetic groups ('phylogroups' hereafter) were chosen to span a wide range ofliving species and to include at least 50 studies, i.e. a number similar to that of published studies in fungi. We eounted the number ofprimer notes for each phylogroup, and for the 50 most recent, we recorded for each polymorphie dinucleotide locus with a minimum of5 repeats, the same items as for the fungal bibliographie data above. A few studies had to be discarded because they reported less than

144 Développement de microsatellites chez les champignons

two polymorphie dinucleotide markers. The complete dataset included 2,923 microsatellite loci.

Data analyses

Using a Mann-Withney's test performed with Statistiea 6.0 (Statsoft 2001), we compared 1) the yield of different steps of microsatellite isolation, 2) the mean repeat number, and 3) the mean allele number per locus and per species, between our dataset and the published studies on other fungi. For the two latter comparisons, only dinucleotides were retained to remove any possible effect ofthe length ofthe motif. To analyse the effect of the phylogroup (i.e. fungi, angiosperms, fishes, insects, birds or mammals) on the mean repeat number and on the mean allele number per species, unpublished studies on fungi were discarded and only polymorphie dinucleotide loci were retained, in order to have data similar to the other phylogroups. The phylogroup effect was tested using an analysis of variance, with the GLM procedure of the SAS software (SAS Institute, SAS Publishing, Cary, NC). Variables were Log-transformed for the residuals to reach normality. Pairwise mean comparisons among phylogroups were performed using Student-Newman-Keuls tests (SNK; Means option in GLM, SAS software). For the mean number of alleles per species, we retained only the studies with a minimum of 14 genotyped individuals to reduce the bias oftoo small a sample size. The effects of the imperfection, the motif (CA/GT and GA/CT, the other dinucleotides being too rare for the analysis), the number of repeats, the sample size and the species on allele numbers were assessed using a generalized linear model (GENMüD procedure of SAS), assuming a Poisson distribution and a log-link function. Because the "allele" variable was over-dispersed, a scaling parameter was calculated to improve the fit to the Poisson distribution. Full models were first fitted including aIl factors and aIl interactions, and then simplified by sequential removal of the least significant highest-order interaction term, retaining signifieant interactions and aIl main effects, even when non-significant.

Results·

Yield ofour 17fungal microsatellite-enriched libraries

145 Annexe 1

In preliminary experiments, we tried to clone fungal microsatellites without enrichment in two species (P. striiformis fsp trUi and P. triticina). The yield was so low (ca. 0.5% of positive clones) that enrichment appeared unavoidable. In our enriched libraries where the clones were screened for the presence of microsatellites, the mean percentage of positive clones (± SE) was 20.2 % (± 5.2). Five libraries had more than 30% of positive clones and four had fewer than 6 %.

After the cloning step, several problems were met due to the method of enrichment. First, a non negligible number ofredundant clones were recovered in aIl experiments, probably due to the two PCR steps required for enrich_ment: the average number of sequences identical to previous sequences had a mean (± SE) of 26.2 % (± 5.1). Second, in three 1aboratories, contamination by foreign DNA occurred in six species and could reach up to 69% of the sequences. These contaminant sequences were easi1y identified: they were repeated severa1 times in the libraries, blasted significantly to sequences in public databases, and/or were sequenced from the previous enrichments performed in the same laboratory. Third, problems in sequencing were met in most of the libraries, in several different laboratories, using either DNA extracted from clones or PCR products purified with various commercial kits. The failure of sequencing reactions seemed to be specific of our adaptors, MluI {Edwards, 1996 #88}, that may adopt a particular 3D structure when linked into vectors from the Topo TA Invitrogen kit, impeding sequencing reactions. The problems in sequencing may also be due to the presence of identical adaptors at each end of the insert. Proper sequences could only be obtained using a particular protocol ofpeR product purification, using PEG {Rosenthal, 1993 #56}. Other studies have used adaptors encompassing a restriction site to avoid this problem {Armour, 1994 #103; Tenzer, 1999 #104}.

In our 17 enriched libraries, the average percentage (± SE) of unique sequences, excluding contaminants, having a microsatellite locus of at least five perfect repeats (Appendix 1) was only 55.4 % (± 4.6) ofthe insert correctly sequenced. Among those, the percentage ofuseful sequences consistently and sharply decreased along the different steps ofthe experiment (Fig. 2). The mean (± SE) of the number of loci eventually polymorphic at the intra-population scale was only of 9.6 % (± 2.5). One of the most critical steps was the suitability of the sequences for primer design (mean ± SE of 56.9 % ± 6.2 of unique sequences with a microsatellite), due to flanking regions with unsuitable base composition or length, or to microsatellites with a too low number of perfect repeats. The percentage of amplified loci

146 Développement de microsatellites chez les champignons

among those suitable for primer design was generaIly high (mean ± SE of 68% ± 8.2), although it was very low in sorne species. In E. alphitoides, E. necator, F poae, M larici­ populina, and P. viticola, only 0 to 45 % ofthe loci retained for primer design could actuaIly be amplified (Appendix 1). The second most important source of attrition was the level of polymorphism obtained from amplifiable loci : the percentage of sequences that eventuaIly yielded polymorphic loci ranged from 0 to 50 % of the initial nurnber of sequenced clones, with only five species above 20% at the inter-population level (Appendix 1) and a mean (± SE) of 17.2 % (± 3.5) (Fig. 2). The intra-population level of polymorphism was even lower, with a mean (± SE) of 9.6 % (± 2.7) of polymorphic loci (Fig. 2). In four species (the two Puccinia spp. and the two Penicillium spp.), no polymorphic loci at aIl could eventually be recovered at the intra-population scale (Appendix 1).

It was not possible to test here if the method of enrichrnent (membrane versus beads) or the length of oligoprobes impacted on the yield of libraries, because the number of studies was too low and because microsatel1ite isolation was performed using different methods in too few species (Appendix 1). However, regardless of the method, the general trend was a poor yield of enriched libraries in fungi. Furtherrnore, in the species for which different methods were used (M larici-populina and E. alphitoides), similar results were observed (data not shown), suggesting a lack of protocol effect. There was also no indication that the libraries enriched for both di- and trinucleotides had a better yield than the libraries enriched only for dinucleotides (means of 17% and 14% of polymorphic loci isolated, respectively). For four out of the ten species enriched with a trinucleotide oligoprobe, no polymorphic trinucleotide loci could be isolated, and for three of them a single polymorphic locus was recovered (data not shown). The mean percentage of polymorphic loci at the largest scale seemed to be slightly higher in the five genomic libraries enriched using only (AC) JO than in those enriched using both (AC/TG)n (means of 25 % and 12% respectively; Appendix 1), but there were too few studies to test this difference given the large species effect (see below).

Comparison between our dataset andthe published studies in fungi

We compared the yield of our enriched libraries with that of the published studies in other fungi to detect possible publication bias or specificities in our data. We col1ected data in the literature from 37 species, among which the proportions of Ascomycota (65%), Basidiomycota (27%) and Oomycota (8%) were similar to our data (Table 1). Among these

147 Annexe 1

37 studies, 43% used beads or membranes to enrich their libraries, 16% the ISSR method, which is based on an amplification using microsatellite primers {Burgess, 2001 #45}, and 27% other methods such as FIASCO {Zane, 2002 #2}, anchored PCR{Zane, 2002 #2}, or search in EST libraries (Appendix 2). Only 14 % of the libraries were not enriched for microsatellites. Few libraries (15%) were enriched only in dinucleotide motifs. Despite the diversity of the methods of microsatellite isolation, the proportion of polymorphie loci relative to the loci tested in fungal species was not significantly different in the literature (mean ± SE of49.7 ± 5.0 % arnong the sequences for which primers were designed, Appendix 2) and in our dataset (53.2 ± 7.4 %; Mann-Withney's test, Z = -0.43, P = 0.67). The mean number of polymorphie loci was higher in our data than in the published studies on fungi (mean ± SE of 14.6 ± 2.9 versus 8.6 ± 0.9, Appendix 2). However, a similar number of loci were observed between our data and the literature when all the published loci were eonsidered, i.e. loci with less than five repeats (c.a. 15 % ofthe total) whieh were not defined as microsatellite in our analysis.

As in our libraries, most of the polymorphie loci in published studies on fungi were dinucleotide (69% and 88% for our data). Considering only the dinucleotide loci, the mean number of repeats per locus and per species was similar in our dataset and in the published studies (11.1 ± 0.7 vs 11.9 ± 0.8; Mann-Withney's test, Z = 0.11, P = 0.92). Within fungi, the mean nurnber ofrepeats per species had a significantly lower mean in Basidiomyeota than in Ascomycota (mean ± SE of 8.8 ± 1.0 versus 12.7 ± 0.7; Mann-Withney's test, Z = 3.2, P = 0.001). In the 22 studies in which libraries were enriched for the two dinucleotides (AC/GT)n and (AG/CT)n, consistently more polymorphic microsatellites were isolated with AC repeats than with AG repeats, regardless ofthe method ofenrichment (5.3 ± 1.3 versus 2.8 ± 0.8 loci per speeies in the published studies and 8.7 ± 2.5 versus 2.3 ± 0.7 in our data, for (AC/GT)n and (AG/CT)n loci, respeetively).

The mean number of alleles per dinucleotide locus per speeies, considering only the studies with a sarnpling size ofat least 14 individuals, was significantly higher in the literature than in our data (5.8 ± 0.4 vs 4.1 ± 0.3; Mann-Withney's test, Z = 2.56, P = 0.01). This result suggests the existence of a publication bias against the microsatellite development work yielding too few polymorphic loci and/or against the less polymorphic loci. When pooling our data and those of the literature, the mean number of alleles was slightly lower in

148 Développement de microsatellites chez les champignons

Basidiomycota than in Ascomycota (mean ± SE of 4.9 ± 0.4 versus 5.4 ± 0.5), but not significantly so (Mann-Withney's test, Z = 0.64, P = 0.52).

Cross-species transferability ofmicrosatellite markers betweenfungal species

Cross-species transferability ofmicrosatellite primer pairs in fungi was estimated based on 24 studies from the literature and our own data. Only 34 % of the 1045 species/primer pair combinations tested within genera were successful in amplifying bands of the expected size. Neither homology, polymorphism nor presence ofnuU aUeles in the transferred microsatellite markers were generaUy assessed.

Comparison offungal microsatellites with those ofother organisms

Only 53% of the published loci were dinucleotides in birds, against ca. 70 % in fungi and fishes and more than 80 % in plants, mammals and insects. More than 34% of the published loci were tetranucleotides in birds, against less than 5% in fungi, insects and plants.

Considering the 46 published studies on fungal microsatellite development with at least two dinucleotide polymorphie loci and the 50 last published Primer Notes on microsatellites isolated in angiosperms, birds, mammals, fishes and insects, the mean number ofdinucleotide polymorphie loci per species was not significantly different among the phylogroups (Kruskal­

Wallis' test, H(5,292) = 8.1, P = 0.15). The mean number ofpolymorphie loci per species was not the lowest in fungi, with a mean ± SE of9.3 ± 1.0, the other phylogroups ranging from 8.5 ± 0.8 (insects) to 13.6 ± 3.4 (fishes).

The number of repeats per dinucleotide locus and per species had a significantly lower mean in fungi (11.8 ± 0.7) than in aU the other phylogroups except birds (13.2 ± 0.6). The means of the other phylogroups ranged from 15.3 (insects) to 17.4 (mammals and fishes) repeats per locus per species (Fig. 3A). The mean number of aUeles per species also had a significantly lower mean in fungi (5.4 ± 0.4) than in aU the other phylogroups (Fig. 3B). AU these phylogroups had similar means of number of aUeles per species, ca. 8 alleles, except the fishes that had a significantly higher mean (11.6 ± 1.2; Fig. 3B).

Factors affecting thediversity ofdinucleotide microsatellites

149 Annexe 1

In fungi, the correlation between the mean repeat number and the mean allele number per species was marginally significantly (r=0.28, P=0.06, Fig. 3). These two variables were significantly correlated in birds (r=0.72, P

A generalized analysis of variance was performed to further investigate which parameters influenced the diversity of individual microsatellite dinucleotide loci in fungi, among motif (considering only AC and AG), imperfection, repeat number, sample size and species (Table 2). The main source of variation affecting the number of alleles of microsatellites was differences among species. A significant effect of the number of repeats was also detected, whereas sample size, imperfection and motifwere not significant.

Discussion

Shared problems were met when isolating microsatellite loci in fungal species. First, the yield of enriched libraries (percentage of positive clones) was low, mostly lower than 30%. This percentage is at the lower limit of what has been obtained in other groups of organisms using the same protocols ofmicrosatellite enrichement, that usually leads to 20% - 90 % ofpositive clones {Zane, 2002 #2}. This may be due to a generallow density ofmicrosatellites in fungal genomes {T6th, 2000 #11; Morgante, 2002 #18} and/or to the rarity of long motifs {Lim, 2004 #5}, which may lower their propensity to hybridize on the probes used for enrichment. Second, the attrition along the different steps from positive clones to polymorphic loci was very high (mean ± SE of 83.8 % ± 3.2). This attrition level is similar to that found in plants, which are known to be recalcitrant species for isolating microsatellite markers {Squirrell, 2003 #84}. For both fungi and plants, the percentage of loci suitable for primer design and the percentage of polymorphic loci seem to be the two most critical steps in microsatellite development. The low percentage of loci suitable for primer design, both in our data and the literature, is partly due to the choice of discarding loci with less than eight perfect repeats. The reason for disregarding them was that very low polymorphism, if any, is generally

150 Développement de microsatellites chez les champignons

expected for such short microsatellites. They are therefore rarely tested when longer loci are available (see e.g. {Lim, 2004 #S}. The positive correlation between number of repeats and number of alleles in fungal species, together with their overall low level of polymorphism, validated a posteriori this choice.

The low occurrence oflong microsatellites recovered in fungal libraries is consistent with the data available in the 14 complete fungal genomes analysed by {Lim, 2004 #S}, where ca. 90% of all microsatellites had between five and seven repeats. Further, the mean percentage of the initial non redundant sequences containing a microsatellite being eventually polymorphic at the intra-population level was less than 10%. In several fungal species, no loci at all could be found polymorphic at the within-population level (Appendix 1), precluding any estimation of population genetics parameters, such·as mode of reproduction or level of gene flow. Finally, a further decrease in the number of usefulloci can occur after this step because of null alleles, physical linkage between loci {Barrès, 200S #S8}, locus duplication, or the existence ofcryptic species which can artificially increase the intra-specific polymorphism for sorne loci {Giraud, 2002 #1O}.

Our data on the 17 enriched libraries showed large discrepancies among speCles III the attrition rate ofusefulloci. In addition to the redundancy ofsequences classically encountered in enriched libraries {Squirrell, 2003 #84}, several of our libraries had high rates of contaminant sequences and/or a low success of primer amplification, in particular in P. viticola, M larici-populina, F poae, E. alphitoides and E. necator. This may be attributed to three specificities of these fungal pathogens: (1) most of them are obligate pathogens that cannot be cultured in vitro as saprophytes, leading to possible contaminations by host plant DNA during microsatellite-enrichment; (2) the existence of cryptic species, widespread in pathogenic fungi {Taylor, 2000 #S9}, are suspected in sorne ofthese species; in such cases, loci cloned in one cryptic species may not be amplifiable in all isolates ofthe morphospecies; (3) sorne of these fungal species may be infected by mycoparasites {Kiss, 1998 #92} or contaminated by bacteria or saprophytic fungi during in vitro culture. The contamination problems may be strengthened in obligate pathogens by the small quantity of DNA that can be obtained after cultivation on their host, which favours the amplification of foreign DNA during the enrichment steps. Our data suggest that contaminations may occur more frequently in microsatellite enriched libraries than previously reported, adding yet another difficulty to microsatellite isolation in fungi. Finally, chimerical inserts generated by the enrichment

151 Annexe 1

protocols {Koblizkova, 1998 #91} cannot be exc1uded. However, the percentage ofamplified loci among those suitab1e for primer design was generally high (mean of 72%) and similar to percentages obtained with primers designed from genomic databases (e.g. {Temnykh, 2001 #78}, suggesting that the enrichment protocol did not generate a high proportion ofchimerical inserts.

AlI these difficulties have certainly played a ro1e in the low number of published microsatellite studies in fungi. For instance, only nine of our 17 microsatellite developments have been published, five of the eight unpublished libraries having yielded each fewer than three loci polymorphie at the intra-population level (Appendix 1). Despite the diversity of methods used to isolate microsatellites in fungi, the published studies on microsatellite development in fungi experienced the same difficulties as our works. The yield of libraries was consistently low, both in terms of quantity and quality of loci isolated, with a few exceptions, which is in agreement with the scarcity of microsatellites in fungal genomes and their shortness (e.g. {Karaoglu, 2005 #4}. These results strongly suggest that several attempts of microsatellite development in other fungal species have probably never been published because of too low a number of polymorphic loci isolated. The proportion of polymorphic loci relative to the number of loci tested and the mean number of alle1es per locus in fungi should therefore be lower than suggested by this survey.

The specifie difficulties in developing fungal microsatellites were further revealed when comparing fungal microsatellites to those of other organisms. The mean number of dinuc1eotide loci isolated in fungi was similar to that of birds or insects, however they were significantly shorter in fungi than in other phylogroups except birds. This difference was even stronger when considering all the published microsatellites, i.e. with fewer than 5 repeats, which were not included in our study according to our microsatellite definition. These very short microsatellites represented almost 15 % of all the published loci in fungi, against less than 5% in the other phylogroups. Despite a preference for loci with the highest numbers of repeats, generally assumed to be the most polymorphic, the scarcity of long microsatellites in the fungal libraries has certainly constrained the choice towards very short microsatellites. The methods used to isolate microsatellites in fungal species may have affected the statistics ofthe cloned loci, in particular their length. However, most ofthe methods used in fungi were also used in the other phylogroups {Zane, 2002 #2}, the enrichments with beads or membranes being prominent in aIl the groups. Despite the noise that the different methods of

152 Développement de microsatellites chez les champignons

isolation may have introduced, the present literature survey based on more than 2200 dinucleotide loci and 250 species showed that microsatellite loci isolated in fungal species had significant1y both fewer repeats and fewer alleles than those in the other phylogroups examined.

Several factors may be responsible for the low polymorphism of microsatellites in fungal populations. Pirst, the lack ofdiversity in fungi is certainly due, at least partly, to the rarity of long microsatellites in these species. Significant correlations between polymorphism of microsatellite loci and their length are known from many other studies. This correlation is explained by the underlying mutation process, i.e. dissociation and mispaired re-association of the nascent strand during DNA replication {Levinson, 1987 #63; Shinde, 2003 #67}. Long microsatellites, with high numbers of perfect repeats, are more likely than short ones to induce DNA replication slippage or unequal crossing-overs {Weber, 1993 #68; Brinkmann, 1998 #69; Goldstein, 1995 #70; Wierdl, 1997 #64; Thuillet, 2002 #79; Schug, 1998 #75; Vigouroux, 2002 #71}. The correlation between microsatellite polymorphism and number of repeats in perfect tandem were confirmed in this study, despite contrasted fungal species and sampling schemes. The effect of microsatellite imperfection on polymorphism was hard to assess with confidence because of the low consistency in the way this feature has been reported in the literature.

So far, we do not have any strongly supported explanation for the low level of repeat number observed in fungal microsatellites. {T6th, 2000 #11} proposed that small genomes, like those of fungi, may possess mechanisms for preventing the accumulation of perfect repeated sequences. However, no correlation was found between genome size and density of repeated sequences when comparing 14 complete fungal genomes {Karaoglu, 2005 #4; Lim, 2004 #5}, although there was a significant positive correlation between genome size and mean length of microsatellites. Measures of mutation rates, great1y facilitated in fungi by their asexual reproduction and rapid generation times, would be valuable for assessing the existence of an efficient mismatching repair system in fungi that limits expansion of microsatellites, as suggested by {Karaoglu, 2005 #4}.

In addition to genomic characteristics, the lower polymorphism observed in fungi compared to other phylogroups may be due to sorne specificities of their life-history traits or history. The variability among fungal species for these characteristics may then explain the large

153 Annexe 1

species effect observed in explaining polymorphism. In particular, many fungal species for which microsatellite markers have been isolated are pathogens (Appendix 2), and many of them have experienced bottlenecks due to these recent introductions or recent host shifts, or selective sweeps due to fungicide selection or resistant crops. Furthermore, most of them regularly undergo asexual reproduction during their epidemic cycles, which can lead to loss of allelic diversity (e. g. {Stephan, 1998 #100}. In our data for instance, the species in which long microsatellites (i.e. higher than 12 perfect repeats in average) have been developed but that exhibited a mean diversity lower than five alleles per locus were generally pathogens on few hosts or were recently introduced (Ascochyta rabei, Cryphonectria parasitica, Magnaporthe grisea, Microcyclus uZei, Mycosphaerella fijiensis, llf. musicoZa, Paecilomyces fumosoroseus, Phytophtora infestans; Appendix 2). However, additional analyses of fungal species are required to be conclusive because the large differences in the numbers and origins of individuals and population history used for the screening of microsatellite diversity preclude detailed analyses for testing for the effect ofthese life history traits.

Too narrow a sampling design may alternatively be invoked to explain the low diversity of microsatellite loci in fungi. However, we recorded the results of fungal diversity from the largest spatial scale tested in each study and the sampling scheme is often chosen to maximise this diversity, by testing the loci on populations as different as possible, i. e. from different countries or continents, or from different host species. There is no evidence that the sample scale in fungi was less representative of the species diversity than that in the other phylogroups.

According to our survey, the overall yield of microsatellite development is low in fungi, indicating that building a genomic library, even with enrichment steps, may be a waste of time and money for many fungal species. However, microsatellites have been easily cloned and have proven very useful in a few species (e.g. {Guérin, 2004 #27; Giraud, 2002 #7; Kaye, 2003 #57}. A good strategy when deciding whether a microsatellite library is worth construeting may therefore be to first evaluate the polymorphism of the speeies using other markers, such as AFLP, ISSR or RAPD. Sueh methods indeed do not require the investment that mierosatellites do for development and may serve as a preliminary test for the level of polymorphism in the focal speeies. If AFLP analysis for instance does not reveal many polymorphie bands, probably microsatellites will not be polymorphie either, and building enriched mierosatellite libraries may just be a waste oftime and money.

154 Développement de microsatellites chez les champignons

If microsateIlite development is to be performed, the present study provides sorne guidelines. First, this survey points to the necessity of using enriched protocols for isolation of microsatellite markers. Our rare attempts to directly clone microsatellites from total DNA failed and 86% of the published microsatellite fungal isolations were performed using enriched protocols or search in databases (Appendix 2). Regarding the methods of enrichment, there was not possible to test their effect on the yield of libraries and the polymorphism of the loci isolated. Our rare attempts comparing the bead and membrane methods on the same species did not suggest any significant differences. Yet, no advice in the choice of the methods can be given, though our experience suggests that enrichment with beads would be more convenient than membranes. Furthermore, it worth noting that the ISSR method, more specificaIly used in fungi, seems to preferentially isolate short microsatellites (Appendix 2). Because this result could be strongly dependent of the species, comparisons of several methods of isolation on the same species are required to provide compelling evidence for such a bias of the ISSR method. Regarding the choice of motif for enrichment, AG appears the most frequent in the published fungal genomes {Lim, 2004 #S}. In contrast, we found that in the studies in which libraries were enriched for the two dinucleotides, consistently more polymorphic microsatellites were isolated with AC repeats than with AG repeats. However, the analyses of fungal genomes also underlined a high variability among species regarding the most frequent motifs. Several genomes were indeed characterized by a low density in the AG motif and appeared richer in AT {Lim, 2004 #S}. The AT motif should nevertheless be avoided as microsatellites with this motif proved to be difficult to amplify, because of the self-complementarity of the motif, and to have low numbers of repeats {Karaoglu, 200S #4}. Because it is therefore difficult to decide a priori which motifs are the most abundant in a given species, the best strategy may be to enrich for both the AC and AG motifs in order to increase the probability to recover numerous and long loci. Finally, although our libraries did not yield many useful trinucleotide markers, enriching libraries in trinucleotides may be worth tryuing. Such motives indeed represented 24% of aIl the polymorphic published loci in fungi .

In addition, our study showed that there is little to expect from cross-species transferability of microsateIlites in fungi. Within genera, only 34% of the loci tested could indeed be transferred, which appears much lower than in animaIs and plants. In plants for instance, 76.4 % of 1800 species/primer combinations tested within genera were successful {Rossetto, 2001

155 Annexe 1

#98}. This discrepancy may be due to higher levels of sequence divergence between fungal species within genera in comparison to animaIs and plants. In fact, the percentage of protein identity between three fungal species belonging to the genus Aspergillus was comparable to that between mammals and fishes (i.e. 66 to 70%; {Galagan, 200S #101}. This lower transferability in fungi compared to other organisms thus suggests that the need to build a specifie library for each species to be studied is even more crucial in fungi. Furthermore, the above estimates of cross-species transferability are certainly over-estimated, because general1y only few individuals per target species have been tested, precluding polymorphism assessment, and no sequencing was carried out to check for the presence ofthe microsatellite locus in the amplified bands.

Once microsatel1ite-enriched libraries have been built, one difficulty is to choose which microsatellites are the best candidates for polymorphie loci. Our analyses confirmed that, after species identity, the best predictor of polymorphism was the length of the longest perfect microsatellite. For developing polymorphie microsatellites, loci with the highest number of perfeet repetitions should therefore be the best candidates. The rule of thumb is to select microsatellites with a minimum of eight repeats, when available {Lim, 2004 #S}. However, we showed that it may be hard in many fungal species to isolate several microsatellite loci with more than eight repeats. Yet, many loci with five to seven repetitions were polymorphie within populations in sorne fungal species, and those may thus be worth testing.

Conclusion

Even if mierosatel1ites loci in fungi seem harder to isolate, to be shorter and to exhibit lower polymorphism than in other organisms, they have been very useful for a number of fungal species. Thus, the huge advantages ofthese markers always render it worth trying to develop as a first approach. However, if the focal species appears to have an overall low genetic diversity and difficulties are met in the first enriched library, more efficient alternatives should rapidly be sought. Among these, AFLP, ISSR and RAPD have the advantage of rapidly screening many loci, but have the drawbacks of being dominant and anonymous. To obtain more reliable and codominant markers for population genetic analyses, specifie primers can be developed to convert anonymous polymorphie bands in sequence eharacterised amplified region (SCAR). Although this method could be as laborious as developing microsatel1ite libraries, it al10ws to obtain a larger number of loci taking

156 Développement de microsatellites chez les champignons

advantage of aIl kind of polymorphisms available in the genome. In addition, the rapid development of genomics may offer a new route for efficiently developing powerful markers in fungi. In particular, expressed sequence tags (EST) or extensive genomic libraries provide an interesting material for finding long microsateIlites or single nucleotide polymorphim (SNPs).

Acknowledgements

This study was funded by the REID (Réseau Ecologie des Interactions Durables), INRA, CNRS and the Université Paris Sud. We thank Jacqui A. Shykoff, Philippe Jarne and Barbara Hefti-Gautschi for helpful discussions and for comments on a previous version of the manuscript and Bruno Le Cam, Marc Seguin and Marc-Henri Lebrun for sharing unpublished data. We also thank the Genoscope, Evry, France for the sequencing of library clones from Magnaporthe grisea, Microcyclus ulei, and Mycosphaerella spp. (project «Plantes et parasites tropicaux - Recherche de locus microsatellites sur plusieurs espèces »). Hans Lawton helped gathering literature data.

References

Armour JA, Neumann R, Gobert S, Jeffreys AJ (1994) Isolation of human simple repeat loci by hybridization selection. Human Molecular Genetics 3,599-565.

Ashley MV, Dow BD (1994) The use of microstellite analysis in population biology: Background, methods and potential applications. In: Molecular Ecology and Evolution, Approaches and Applications (eds. Schierwater B, Streit B, Wagner G, DeSalle R), pp. 185­ 201. Birkauser Verlag, Basel .

Barrès B, Dutech C, Andrieux A, et al. (2006) Isolation and characterization of 15 microsatellite loci in the poplar rust fungus, Melampsora larici-populina, and cross­ amplification in related species. Molecular Ecology Notes 6: 60-64.

Breuillin F, Dutech C, Robin C (2006) Genetic diversity of the chestnut blight fungus Cryphonectria parasitica in four French populations assessed by microsatellite markers. Mycological Research.11 0:288-296.

157 Annexe 1

Brinkmann B, Klintschar M, Neuhuber F, Huhne J, Rolf B (1998) Mutation Rate in Human Microsatellites: Influence of the Structure and Length of the Tandem Repeat. American Journal ofHuman Genetics 62, 1408-1405.

Bucheli E, Gautschi B, Shykoff JA (2000) Host specific differentiation in the anther smut fungus Microbotryum violaceum as revealed by microsatellites. Journal of Evolutionary Biology 13, 188-198.

Burgess T, Wingfield M, Wingfield B (2001) Simple sequence repeat markers distinguish among morphotypes of Sphaeropsis sapinea. Applied and Environmental Microbiology 67, 354-362.

Cornuet J-M, Piry S, Luikart G, Estoup A, Solignac M (1999) New methods employing multilocus genotypes to select or exclude populations as origins of individuals. Genetics 153, 1989-2000.

Delmotte F, Chen W-J, Richard-Cervera S, et al. (2006) Microsatellite loci from the grape downy mildew (Plasmopara viticola). Molecular Ecology Notes in press.

Dettman JR, Taylor JW (2004) Mutation and evolution of microsatellite loci in Neurospora. Genetics 168, 1231-1248.

Dieringer D, Schlotterer C (2003) Two distinct modes of microsatellite mutation processes: Evidence from the complete genomic sequences ofnine species. Genome Research 13, 2242­ 2251.

Duan X, Enjalbert J, Vautrin D, Solignac M, Giraud T (2003) Isolation of twelvemicrosatellite loci, using an enrichment protocol, in the phytopathogenic fungus Puccinia triticina. Molecular Ecology Notes 3,65-67.

Edwards KJ, Barker JHA, Daly A, Jones C, Karp A (1996) Microsatellite librairies enriched for several microsatellite sequences in plants. Biotechniques 20, 758-760.

158 Développement de microsatellites chez les champignons

Ellegren H (2004) Microsatellites: Simple sequences with complex evolution. Nature Review Genetics 5, 435-445.

Engelbrecht C, Harrington T, Steimel J, Capretti P (2004) Genetic variation in eastem North American and putatively introduced populations of Ceratocystis fimbriata f platani. Molecular Ecology 13, 2995-3005.

Enjalbert J, Duan X, Leconte M, Hovmoller M, De Vallavieille-Pope C (2005 ) Genetic evidence of local adaptation of wheat yellow rust (Puccinia striiformis f sp tritici) within France Molecular Ecology 14 2065-2073

Enjalbert J, Duan X, Vautrin D, et al. (2002) Isolation of twelve microsatellite loci, using an enrichment protocol, in the phytopathogenic fungus Puccinia striifomisfsp. tritici. Molecular Ec%gy Notes 2, 563-565.

Field D, Wills C (1998) Abundant microsatellite polymorphism in Saccharomyces cerevisiae, and the different distributions of microsatellites in eight prokaryotes and S. cerevisiae, result from strong mutation pressures and a variety of selective forces. Proceedings olthe National Academy ofSciences ofthe USA 95, 1647-1652.

Fisher P, Gardner R, Richardson T (1996) Single locus microsatellites isolated using 5' anchored PCR Nucleic Acids Research 24 4369-4371.

Galagan JE, al. e (2005) Sequencing ofAspergillus nidu/ans and comparative analysis with A. fumigatus andA. oryzae. Nature 438, 1105-1115.

Giraud T, Fournier E, Vautrin D, Solignac M, Shykoff JA (2002a) Isolation of 44 polymorphie microsatellite loci lU three host races of the phytopathogenic fungus Microbotryum vio/aceum. Mo/ecu/ar Ec%gy Notes 2, 142-146.

Giraud T, Fournier E, Vautrin D, et al. (2002b) Isolation of eight polymorphie microsatellite loci, using an enrichment protocol, in the phythopathogenic fungus Fusarium culmorum. Mo/ecu/ar Ec%gy Notes 2, 121-123.

159 Annexe 1

Goldstein D, Clark A (1995) Microsatellite variation in North American populations of Drosophila melanogaster. Nucleic Acids Research 23, 3882-3886.

Goodwin S, Cohen B, Fry W (1994) Panglobal distribution of a single clonallineage of the Irish potato famine fungus. Proceedings ofthe National Academy ofSciences ofthe USA 91, 11591-11595.

Guérin F, Le Cam B (2004) Breakdown of the scab resistance gene Vf in apple leads to a founder effect in populations of the fungal pathogen Venturia inequalis. Phytopathology 94, 364-369.

Hardy OJ, Charbonnel N, Fréville, H, Heuertz M (2003) Microsatellite allele sizes: A simple test to assess their significance on genetic differentiation. Genetics 163, 1467-1482.

Hawksworth DL (2001) The magnitude of fungal diversity: the 1.5 million species estimate revisited. Mycological Research. 105, 1422-1432.

Hovm011er MS, Justesen AF, Brown JKM (2002) Clonality and long-distance migration of Puccinia striiformis f. sp. tritici in north-west Europe. Plant Pathology 51, 24-32.

Jarne P, Lagoda PJL (1996) Microsatellites, from molecules to populations and back. Trends in Ecology and Evolution 11,424-429.

Karaoglu H, Lee CMY, Meyer W (2005) Survey of simple sequence repeats in completed fungal genomes. Molecular Biology and Evolution 22, 1159-1159.

Kaye C, Milazzo J, Rozenfeld S, Lebrun M, Tharreau D (2003) The development of simple sequence repeat markers for Magnaporthe grisea and their integration into an established genetic linkage map. Fungal Genetics andBiology 40, 207-214.

Kijas JMH, Fowler JCS, Garbett CA, Thomas MR (1994) Enrichment ofmicrosatellites from the Citrus genome using biotinylated oligonucleotide sequences bound to streptavidin-coated magnetic particles. Biotechniques 16, 656-660.

160 Développement de microsatellites chez les champignons

Kiss L (1998) Natural occurrence ofAmpelomyces intracellular mycoparasites in mycelia of powdery mildiew fungi. New Phytologist 140,709-714.

Koblizk:ova A, Dolezel J, Macas J (1998) Subtraction with 3' modified oligonucleotides eliminates amplification artifacts in DNA libraries emiched for microsatellites. Biotechniques 25,32-38.

Le Guen V, Rodier-Goud M, Troispoux V, et al. (2004) Characterization of polymorphic microsatellite markers for Microcyclus ulei, causal agent of South American leaf blight of rubber trees. Molecular Ecology Notes 4, 122-124.

Levinson G, Gutman GA (1987) Slipped-strand mispairing: a major mechanism for DNA sequence evolution. Molecular Biology and Evolution 4, 203-221.

Lim S, Notley-McRobb L, Lim M, Carter DA (2004) A comparison of the nature and abundance ofmicrosatel1ites in 14 fungal genomes. Fungal Genetics and Biology 41,1025-36.

Luikart G, England PR, Tallmon D, Jordan S, Taberlet P (2003) The power and promise of population genomics: from genotyping to genome typing. Nature Reviews Genetics 4, 981­ 994.

Mackenzie JS, Chua KB, Daniels W, et al. (2001) Emerging viral diseases of Southeast Asia and the western Pacific. Emerging Infectious Disease 7,497-504.

Michalakis Y, Excoffier L (1996) A generic estimation of population subdivision using distances between alleles with special interest to microsatellite loci. Genetics 142, 1061-1064.

Milgroom MG, Liparis S, Wang KR (1992) Comparison of genetic diversity in the chestnut blight fungus Cryphonectria (Endothia) parasitica, from China and the United States. Mycological Research 96, 1114-1120.

Molina C, Kaemmer D, Aponte S, Weising K, Kahl G (2001) Microsatellite markers for the fungal banana pathogen Mycosphaerella musicola. Molecular Ecology Notes 1, 137-139.

161 Annexe 1

Morgante M, Hanafey M, Powell W (2002) Microsatellites are preferentially associated with non repetitive DNA in plant genomes. Nature Genetics 30, 194-200.

Nei M (1987) Molecular Evolutionary Genetics Columbia University Press, New York.

Neu C, Kaemmer D, Kahl G, Fischer D, Weising K (1999) Polymorphie microsatellite markers for the banana pathogen Mycosphaerellafijiensis. Molecular Ecology 8,523-525.

Paetkau D, Slade R, Burden M, Estoup A (2004) Genetic assignrnent methods for the direct, real-time estimation of migration rate: a simulation-based exploration ofaccuracy and power. Molecular Ecology Notes 13, 55-65.

Paraskevis D, Lemey P, Salemi M, et al. (2003) Analysis of the evolutionary relationships of HIV-1 and SIVcpz sequences using Bayesian inference: Implications for the origin of HIV-1. Molecular Biology and Evolution 20, 1986-1996.

Petit, RJ, Deguilloux M, et al. (2005) Standardizing for microsatellite length in comparisons ofgenetic diversity. Molecular Ecology 14, 885-890.

Rivas G, Zapater M, Abadie C, Carlier J (2004) Founder effects and stochastic dispersal at the continental scale of the fungal pathogen of bananas Mycosphaerella fTjiensis. Molecular Ecology 13, 471-482.

Rosenthal A, Coutelle 0, Craxton M (1993) Large-scale production of DNA sequencing templates by microtitre format PCR. Nucleic Acids Research 21, 173-174.

Rossetto M (2001) Sourcing of SSR markers from related plant species. In plant genotyping : the DNA fingerprinting of plants (ed. Henry RJ), pp. 211-224. CABI Publishers, New York, N.Y.

Schug MD, Hutter CM, Wetterstrand KA, et al. (1998) The mutation rates of di-, tri- and tetranucleotide repeats in Drosophila melanogaster. Molecular Biology and Evolution 15, 1751-1760.

162 Développement de microsatellites chez les champignons

Shinde D, Lai YL, Sun FZ, Arnheim N (2003) Taq DNA polymerase slippage mutation rates measured by PCR and quasi-likelihood analysis: (CA/GT)(n) and (AlT)(n) microsatellites. Nucleic Acids Research 31, 974-980.

Squirrell J, Hollingsworth PM, Woodhead M, et al. (2003) How much effort is required to isolate nuclear microsatellites from plants? Mol Eco112, 1339-1348.

Stephan W, Langley CH (1998) DNA polymorphism in Lycopersicon and crossing-over per physicallength. Genetics 150, 1585-1593.

Symonds VV, Lloyd AM (2003) An analysis of microsatellite loci in Arabidopsis thaliana: mutational dynamics and application. Genetics 165, 1475-1488.

Taylor JW, Jacobson DJ, Kroken S, et al. (2000) Phylogenetic species recognition and species concepts in Fungi. Fungal Genetics and Biology 31,21-32.

Temnykh S, DeClerck G, Lukashova A, et al. (2001) Computational and experimental analysis of microsatellites in rice (Oryza sativa L.): Frequency, length variation, transposon associations, and genetic marker potential. Genome Research 11, 1441-1452.

Tenzer I, degli Ivanissevich S, Morgante M, Gessler C (1999). Identification ofmicrosatellite markers and their application to population genetics of Venturia inaequalis. Phytopathology 89, 748-753.

Thuillet AC, Bru D, David J, et al. (2002) Direct estimation of mutation rate for 10 microsatellite loci in dururn wheat, Triticum turgidum (L.) TheIl. ssp dururn desf. Molecular Biology and Evolution 19, 122-125.

Tobler K, Ackermann M, Griot C (2003) SARS-agent and lessons to be learned from pathogenic coronaviruses ofanimaIs. Schweizer Archivfür Tierheilkunde 145, 316-322.

T6th G, Gaspari Z, Jurka J (2000) Microsatellites in Different Eukaryotic Genomes: Survey and Analysis. Genome Research 10, 967 - 981.

163 Annexe 1

Tyler BM (2001) Genetics and genomics of the oomycete-host interface. Trends in Genetics 17,611-614.

Vigouroux Y, Jaqueth JS, Matsuoka Y, et al. (2002) Rate and pattern of mutation at microsatellite loci in maize. Molecular Biology and Evolution 19, 1251-1260.

Watterson GA (1975) On the number of segregating sites ln genetical models without recombination. Theoretical Population Biology 7,256-276.

Weber J, Wang C (1993) Mutation of human short tandem repeats. Human .Molecular Genetics 2, 1123-1128.

Wierdl M, Dominska M, Petes TD (1997) Microsatellite instability in yeast: dependence on the length ofthe microsatellite. Genetics 146, 769-779.

Zane L, Bargelloni L, Patarnello T (2002) Strategies for microsatellite isolation: a review. Molecular Ecology 11, 1-16.

164 Développement de microsatellites chez les champignons

Figure legends

Figure 1. Number of primer notes published in Molecular Ecology Notes between March 2001 and June 2005 (black bars) on fungi, angiosperms, fishes, birds, mammals and insects. For fungi, microsatellite isolation reports found in other joumals are represented in grey.

Figure 2. Evolution ofthe average percentage ofuseful sequences along the different steps for 17 fungal microsatellite-enriched libraries. Percentages are given for each step relative to the initial number ofinserts correctly sequenced. Maximum and minimum values are given by the verticallines and dashed lines indicate limits ofthe first and the third quartiles. See Appendix 1 for details on the libraries.

Figure 3. Boxplots of the number of repeats (A), and of the number of alleles (B) for fungi, angiosperms, fishes, birds, mammals and insects. Statistics are represented on the means per species for aIl primer notes published in Molecular Ecology Notes between March 2001 and June 2006 (and published in all joumals for fungi). Boxes indicate quartiles, dark squares means and vertical traits minimal and maximal values. Different letters indicate significantly different groups ofmeans in a SNK pairwise comparison test (p

Figure 4. Mean number ofalleles per species plotted against mean number ofrepeats (number of repeats of the longest pure tandem repeat) for aIl available data on fungal microsatellites. The regression line is drawn (r=0.28, P=0.06).

165 Annexe 1

Table 1. Fungal species for which the authors built microsatellite-enriched libraries, with their characteristics.

Number of Method of Species Order Motiffor enriched libraries screened enrichment individuais Cryphonectria parasitica Ascomycota membrane (AC)15 (AG)15 (AAG)IO 113

Erysiphe alphitoides Membrane (ACb (AG)15 (AAG)IO Ascomycota 8 beads (AC)15 (AG)15

Erysiphe necator Ascomycota beads (TC) 10 (TG) 10 15

Fusarium culmorum Ascomycota beads (AC)IO (AAG)IO 20

Fusarium poe Ascomycota beads (AC)10 20

Magnaporthe grisea Ascomycota beads (TC) 15 (AC) 15 6

Melampsora larici- membrane (AC) 15 (AG) 15 (AAG) 10 populina Basidiomycota beads (AC) 15 (AG) 15 30 (5 libraries) beads (TC) 10 (TG) 10

Microbotryum violaceum Basidiomycota beads (AC)IO (AAG)IO 30 (3 1ibraries)

15 Mycosphaerella musicola Ascomycota beads (TC) 15 (AC) 15

Mycosphaerella fyiensis Ascomycota beads (TC) 15 (AC)15 15

Mycosphaerella eumusae Ascomycota beads (TC) 15 (AC) 15 15

Mycrocyclus ulei Ascomycota beads (TC) 15 (AC) 15 16

Penicillium roqueforti Ascomycota beads (AC)IO (AAG)IO (CAC)IO (GGA)10 5 (4libraries)

Penicillium camembertii Ascomycota beads (AC)IO (AAG)IO 5

Plasmopara viticola Oomycota beads (TC) 10 (TG) 10 (GAA) 10 (TAA) 10 100

Puccinia striiformisfsp Basidiomycota beads (AC) 10 (AG) 10 (AAC) 10 (AAG) 10 96 tritici

Puccinia triticina Basidiomycota beads (AC) 10 (AG) 10 (AAC) 10 (AAG) 10 15

166 Développement de microsatellites chez les champignons

Table 2. Results ofthe GENMOD analysis testing for an effect ofrepeat number, imperfection and motifofthe loci, species and sample size on the allele number ofmicrosatellites (R2=0.49).

Source D.F. Chi Square P Number ofrepeats 1 25.34 <.0001 Imperfection 1 1.59 0.2076 motif 1 0.48 0.4888 Species 40 223.39 <.0001 Sample size 1 0.49 0.4828

167 Annexe 1

Number of published papers

140

120

100

80

60

40

20

o Fungi Angiosperms Fishes Birds Mammals Insects '------...... ------_../ Animais Phylogroups

Dutech et al. Figure 1

168 Développement de microsatellites chez les champignons

...... III .... III!! T l .. '" "1.... T------.:JL-.J .. Il .. III .... III l ...... - .. Ils .. III .. .. III ...... III .. ..

Sequences with Tested loci Amplifiable loci Polymorphie loci Polymorphie loci mierosatellites among within populations populations

Dutech et al. Figure 2

169 Annexe 1

Number of repeats

50 r----,,------r------.----.,.----,.------,----,---..,.---, c 45

40 bc 35 ab 30 a bc c 25 T -r 1 20

15 ~ ';;;::;;::::;1::;;:;:,;:;

10

5

oL------'----'-----'------'-----'-----'------'-----'-----l Fungi Birds Insects Angiosperms Fishes Mammals

Figure 3A Dutech et al.

170 Développement de microsatellites chez les champignons

Number of aileles

50 r---r---.------r------.----r----r------,~--_.____,

40 c

30 bc bc bc b T 20 TT a

10

0'----'----"-----&..--,...----''-----'------'----''-----&..---1 Fungi Birds Insects Mammals Angiosperms Fishes

Figure 3B Dutech et al.

171 Annexe 1

Number of aleIles

16

14

12

10

6

4

2

o-+------.,.----~----,----~---~ o 5 10 15 20 25

Number of repeats

Outech et al. Figure 4

172 Développement de microsatellites chez les champùmons

Appendix 1. Proportion ofusefulloci along the different steps ofrnicrosatellite developrnent in 17 fungal species. The percentages are relative ta the initial nurnber ofsequenced clones.

Percentage ofunique Percentage of Percentage ofsequences Percentage ofsequences Numberof Percentage ofsequences Species sequences witb a sequences suitable for yielding correct yielding polymorphie yielding intra-population sequenced clones microsatellite primer design amplification loci at the largest scale polymorphie loci

Cryphonectria parasitica 58 44.8 29.3 27.6 17.2 17.2

Fusarium culmorum 33 57.6 27.3 27.3 24.2 23.7

Fusarium poae 137 63.5 14.6 1.5 1.5 1.5

Erysiphe alphitoides 110 19.1 9.1 0.0 0.0 0.0

Erysiphe necator 96 79.2 18.8 8.3 3.1 2.1

Magnaporthe grisea 96 46.9 20.8 10.4 10.4 10.4

Melampsora larici-populina 208 17.8 17.8 7.2 6.7 6.3

Microbotryum violaceum 143 42.0 30.8 30.8 30.8 13.0

Microcyclus ulei 52 46.2 26.9 25.0 21.2 21.2

Mycosphaerella eumusae 58 63.8 32.8 19.0 15.5 15.5

Mycosphaerella fijiensis 111 46.8 43.2 29.7 29.7 29.7

Mycosphaerella musicola 64 53.1 34.4 18.8 18.8 18.8

Penicillium camembertii 24 66.7 54.2 50.0 50.0 0.0

Penicillium roqueforti 23 43.5 43.5 39.1 17.4 0.0

Plasmopara viticola 186 37.1 19.9 4.8 3.8 3.8

Puccinia striiformisfsp 79 75.9 35.4 30.4 15.2 0.0 tritici Puccinia triticina 130 82.3 27.7 23.8 9.2 0.0

9.6 ± 2.5 Mean 52.1 ±4.7 28.6 ± 3.9 20.8 ± 3.5 l6.2±3.2

173 Annexe 1

Appendix 2. Isolation and characterization ofrnicrosatellite loci in fungal species

OUR DATA

Species Method oflibrary Nb oftested Percentage of Percentage of Nb ofloci used Nb of Nb of Nb ofstrains References enrichment primers scorable loci polymorphie loci for analyses(l) repeats(2) alleles(2) tested

Membrane and Cryphonectria parasitica EST(3) 17 94 59 8 13.3 4.9 113 Breuillin et al. (2006)

Erysiphe necator Beads 18 44 11 DA DA DA DA Unpublished

Fusarium culmorum Beads 9 100 89 7 13.0 6.1 20 Giraud et al. (2002a)

Fusarium poae Beads 20 10 10 DA DA DA DA Unpublished

Magnaporthe grisea Beads and EST() 19 50 30 24 16.7 2.9 6 Kaye et al. (2003) and unpublished

Melampsora lariei-populina Beads and membrane 37 41 35 12 9.7 4.9 30 Barrès et al. (2005)

Microbotryum violaceum Beads 44 100 100 44 11.4 5.4 30 Giraud et al. (2002b)

Microcyclus ulei Beads 14 93 93 II 15.0 2.9 16 Le Guen et al. (2004) and unpublished

Mycosphaerella eumusae Beads 19 58 47 9 8.6 2.9 15 Unpublished

Mycosphaerella fijiensis Beads 48 69 69 33 10.2 4.6 15 Unpublished

Mycosphaerella musicola Beads 19 55 55 12 10.2 4.1 15 Unpublished

Unpublished Penicillium camembertii Beads 13 100 0 Il 12.2 2.6 5

Unpublished Penicillium roqueforli Beads 19 84 0 4 11.8 3.5 5

4.5 100 Delmotte et al. (2006) Plasmopara viticola Beads 15 7 7 6 8.2

2.5 96 Enjalbert et al. (2002) Puccinia striiformisftp trWei Beads 28 86 43 Il 7.3

8.1 3.2 15 Duan et al. (2003) Puccinia triticina Beads 36 86 33 12

14.6±2.9 1l.l±0.7 4.1 ± 0.3(4) 44.4 ± 12.1 Mean 23.4 ± 3.0 69.9± 7.0 53.2 ± 7.4

174 Développement de microsatellites chez les champifmons

LITERATURE DATA

Species Method of library Nb oftested Percentage of Percentage of Nb ofpublished Nb of Nb ofallelesw Nb ofstrains References enrichment primers scorable loci polymorphic loci loci (1) repeats(2) tested

Alternaria brassicicola Beads 21 81 57 12 8.7 2.9 46 Avenot et al. (2005)

Armillaria ostoyae Membrane 12 100 100 12 9.9 7.2 23 Langrell et al. (2001)

Ascochyta rabiei Non enriched 37 70 54 20 12.6 4.2 22 Geistlinger et al. (2000)

Aspergillusflavus GeneBanlP) NA NA NA 6 13.4 7.4 20 Tran-Dinh & Carter (2000)

Aspergillusparasiticus GeneBank NA NA NA 6 13.4 5.2 15 Tran-Dinh & Carter (2000)

Aspergillusfumigatus Non enriched 8 100 50 4 19.3 14.0 100 Bart-Delabesse et al. (1998)

Beauveria brongniartii Beads 11 91 91 10 11.8 3.8 96 Enkerli et al. (2001)

Beauveria bassiana Beads NR NR NR 8 20.0 6.3 24 Rehner & Buckley (2003) Botryosphaeria parva ISSR(6) 20 NR 35 5 9.0 3.0 82 Slippers et al. (2004)

Botryosphaeria rhodina ISSR 19 68 42 5 5.0 4.0 9 Burgess et al. (2003)

Coccidioides immitis Non enriched 9 100 100 9 \3.6 9.9 25 Fisher et al. (1999)

Cryphonectria eucalypti FIASCO(7) NR NR NR 8 20.3 7.0 20 Nakabonge et al. (2005)

Cylindrocladium parasiticum ISSR 25 NR 48 8 7.5 3.5 17 Wright et al. (2006)

Dothistroma pini Anchored and EST 9 78 33 4 6.7 2.7 12 Ganley & Bradshaw (2001)

Fusarium oxysporum rSSR NR NR NR 9 12.5 Il 64 Bogale et al. (2005) Johannesson & Stenlid (2004) Heterobasidion annosum Beads NR NR NR 5 1l.3 6.5 19 Jany et al. (2006) Laccaria bicolor Beads 150 26 5 7 8.7 7.0 12 14 Hayden et al. (2004) Leptosphaeria maculans Non enriched and EST 36 94 19 6 9.5 2.3 30 Walser et al. (2003) Lobaria pulmonaria Beads 21 57 52 Il 14.5 8.3 102 Barett & Brubaker (2006) Melampsora lini FIASCO 41 83 27 10 7.5 4 5.4 34 Enkerli et al. (2005) Metarhizium anisopliae Beads 14 100 100 8 10.6 20.2 5.0 29 Neu et al. (1999) Mycosphaerella fijiensis Beads 19 58 58 10 22.1 3.4 24 Molina et al. (2001) Mycosphaerella musicola Beads 48 NR 54 24 11 5 77 Feau et al. (2006) Mycosphaerella populorum EST 10 NR 50 3 14.4 3.6 26 Dalleau-Clouet et al. (2005) Paecilomycesfùmosoroseus Beads 9 100 89 8 6.0 6.0 121 Stukenbrock et al. (2005) Phaeosphaeria nodorum EST 73 NR 12 11 20.5 5.3 5 Dobrowolski et al. (2002) Phytophthora cinnamomi Membrane 10 40 40 4 175 Annexe 1

Phytophthora ramorum Beads 24 58 29 7 10 3.5 26 Prospero et al. (2004)

Phytophthora infestans Non enriched and EST 108 NR Il 13 15.3 4.7 90 Lees et al. 2006

Pisolithus microcarpus RAPD-PCR 25 NR 24 4 5.5 4.0 30 Hitchcock et al. 2006

Pisolithus spp. ISSR 8 100 75 3 5.0 4.5 39 Kanchanaprayudh et al. (2002) Pleurotus eryngii Beads 6 83 83 5 10.0 5.0 22 Della Rosa et al. (2004)

Russula brevipes Enriched 26 23 23 5 6.3 5.7 27 Bergemann et al. (2005)

Sclerotinia sclerotiorum Membrane 70 NR 36 17 9.9 3.7 44 Sirjusingh & Kohn (2001)

Sphaeropsis sapinea ISSR 22 82 68 3 8.5 8.0 40 Burgess et al. (2001)

Tuber magnatum Beads 23 91 35 8 18.5 7.5 370 Rubini et al. (2004)

Venturia inaequalis Beads 52 40 40 17 10.7 JO. 1 44 Guérin et al. (2004)

Mean 3I.2±5.8 75.0 ± 5.3 49.7 ± 5.0 8.6 ± 0.8 11.9 ± 0.8 5.8 ± 0.4(4) 52.2 ± 10.8

NA: Not applicable NR: Not reported (1) Number ofpolymorphic loci with a minimum offive repeats (2) Only dinucleotide loci considered (3) Loci isolated from an expressed sequence tags (EST) library (4) Loci isolated from genomic data available at http://www.ncbi.nlm.nih.gov/ (3) Loci isolated from an enriched library using an intergenic simple sequence repeats (ISSR) protocol {Burgess, 2001 #45} (4) Loci isolated from an enriched library using an anchored protocol {Fisher, 1996 #95} (5)

(6) Loci isolated from an enriched library using a FIASCO protocol {Zane, 2002 #2} (7) Mean estimated using only the studies with at least 14 haplotypes.

References

176 Annexe 2 Développement de microsatellites chez Phytophthora alni spp. alni

Molecular Ecology Notes (2007) 7, 133·-137 doi: 10.1111/j.H71-8286.2006.01554.x

PRIMER NOTE Characterization of microsatellite markers in the interspecific hyhrid Phytophthora alni ssp. alni, and cross-amplification with related taxa

RENAUD IOOS,*t BENOÎT BARRÈS: AXELLE ANDRIEUX* and PASCAL FREY* *liVRA, Equipe de Pathologie Forestiere, UMR 1136 Interactions Arbres ---lv1.icroorganismes, IFR 110,54280 Champenoux, France, tl.nboratoire National de la Prote,ction des VégétatLT, UMAF, Domaine de Pixérécourt, 54220 Malzélnlle, Franœ

Abstract Phytophthora alni ssp. alni is an interspecific hybrid oomycete causing a large-scale decay of aMers throughoutEurope. In this study we developed a set of 10 microsatellite markers that shows p.romise for population studies and for studying hybridization events between the parental species of the hybrid. Moreover, the genotype and the ploidy of the different subspecies ofF. al-rd might be inferred from the quantitative ratio of amplified genome­ specifie alleles. Nine primer pairs cross amplified with the related species Phytophthora c.ambivora and Phytophthorafragariae and yielded distinct alleles. Keywords: genotyping, oomycete, polyploidy

Received 29 June 2006; reI'ision 7 Au3"'Ust 2006

Phytophthora alni ssp. alni (Paa) is a reœntly described inter­ and, the two other taxa, we isolated and characterized specifie hybrid oomycete causing a lethal disease of aIder mierosatellite loci in Paa. These codominant markers are throughout Europe (Brasieret al. 2004). Two other taxa close especially useful to address questions relating to origin to Paa are isolated from diseased alders or from soi! beneath and genetic diversity of fungi (Weising et al. 1995). alders: Phytapht/wYa alni ssp.unifonnis (Pau) and Phytapht/wYa Microsatel1ite markers were isolated from Paa t1Sing alni ssp. multiforrnis (Pam). Ail threetaxa are phylogenetically enrichment libraries. DNA was extracted from Paa isolate very close ta Phytophthora çambivora (pc) and Phytophthora PA.A130ush'g DNeasyPlantMini Kit (QIAGEN). Approx­ fmgariae (Pf) (Brasier et al. 1999). Previously hypothesiZed ta imately 500 ng f.LL -1 DNA were recovered and used for the originate from geneticbreakdown. ofPaa (Brasier el al. 1999), construction of Cive enJ'iclunent libraIies using biotinylated Pam and Pau were actuaUy suggested to represent distinct oligoprobes [(AC)lS' (AG)lS' (ACG)6' (AAC)6 and (AAG)6] species and to have generated Paa through interspecific and streptavidin-eoated magneticbeads, accordingto Dutech hybridization (Iooseta.l. 2006). In addition, Brasier et al. et al. (2000). The libraries were cloned using a TOPO TA (1999)reported thatPaawas anear tetraploid species (4n + Cloning kit (Invitrogen). For each ofthe five enrichments, 2) in respect with the typical diploid (2n) status of the 288Clones werescreenedbycolonyblotfoUowing a protocol Phytophthora genus, and iliat Pamand Pau Were aneuploid deve10ped by Estoup et al. (1993). A total of 172 positive species with 2n + 4 to 2n + 7, .and 2n + 2 chromosomes, clones were sl1bsequently size-screened by polymerase respectively. However, loos et al. (2006) demonstrated that chain reaction (PCR) lISing vector's primers and 43 clon.es three divergent alleles for four unlinked single copy nuclear with insert between 400 and 1000 bp were retained for genes were observed withù1 the Paa genome. Two of{hem sequencing. Eighteenprimer pairscould be designed from matched the two alleles also foundin Pamand the third tl1e sequences,either manually or using PRIMER3 software one corresponded tothe singleallele found in Pau. As these (Ro.zen & Skaletski 2000). The primer pairs were tested by results contrasted with the ploidy levels reported, and as PCR with a panel of6 Paa, 'iPam, 8 Pau, 2 Pc and 2 Pl isolates, little is known about the genetic polymorphismwithin Paa from various geographical origins in Europe and displaying different mitotypes according to loos et al. (2006) (Table 1). Correspondence: P. Frey, Fax: +33383394069, PCR conditions were optimized for each primer pair. PCR E-rnail: [email protected] was conducted in a 20-f.LL volume containing 15-20 ng of

© 2006 The Authors Journal compilation © 2006B1ackwell Pnblishing Ltd

177 Annexe 2

134 PRIMER NOTE

Table 1 List of the isolates used in the study. Mitotypes oHhe isolates detennined as described in loos et aL (2006)

Geographical Year Species lsolate Mitotype Host origin isolated

Phytnphthora a/ni ssp. a/ni PAA129 U Aln.us glutinosa SW France 2003 PAA130' M Alnus glutinosa NE France 2003 l'AA143 1\.1'" A/nus g/utinClSa l'oland 2002 PAA151 V Alnus glutinosa NE France 2004 PAA162 II Alnus glulinosll Germany 2004 PAAa2' V Alnus glutinosa NE France 2005 PAAe3 M A/nus glutinosll NE France 2005 pj\A151 U A/nus glutin osa NE France 2004 PAA70 M Alnus g/utinosa the Netherlands Onknown PA..<\a2 M Alnus g/utinosa NE France 2005 l'AA38 M Alnus g/utinosa N France 2002 l'A/\.74 U Alnus glutinosa Scotland 2000 PAA75 M A/nus vtridis Scotland 2000 l'Ao\.81 U Alnus glutinosa England 1997 PA<\82 M A/nus g/utinosa England 1996 PAAel M A/nus g/utÎnosa NE France 2005 PAA85 M A/nus: g/1ltinosa England Unknown l'AA86 M Ahzus g/utinosa Belgium 1999 PA..<\88 M A/nus g/utinosa Belgium 2001 PAA91 M' A/nus g/1ltinosa H\lllgary 2001 PAld12 M Alnus g/utinosa N'.,V Fra nce 2003 PAA114 M Alnus g/utinosa NE France 2002 PA<\134 M Alnus glutinosa Germany 2000 PAA141 U' A/nus glutinosa Austriil Unknown PAi\.144 M" Alnus glutin osa Poland 2003 PAA178 M" Alnus g/utinosa l'oland 2003 PAA185 M Alnus glutinosa NE France 2004 PM194 M Alnus gluti1losa NWFrance 2005 PAAd6' U Alnusglatinosa NW France 2005 l'MalO M Alnus g/utinClSa N France 2005 l'MelO M Alnus g/utinosa N France 2005 P.'V\d2' M' Alnus glutinosa NE France 2005 PAAé2' M Alnus glutinosa NEFrance 2005 l'AAall' M Alnus g/utinosa Gem1any 2005 Phytophtht-.ra a/ni ssp. mu/tiformis P.AM54 M A/nus g/utin&'Sa N1N France 2000 PAt"I7! M Alnus: glutinosasoil the Nethi'Ilands 1995 l'A..I\I90 M A/nus g/utinosa soil the Netherlands 1995 PAM186 M' A/nus g/utinosa Belgium 2001 l'M173 nd Alnus g/utinosa England 1996 Phytophthora alni ssp,uniforinis PAU60 U" A/nus glutinosa NE France 1999 PAU84 V' Alnus g/utinosa Swéden 1999 l'AU87 U A/nus g/utinosa Bëlgium 2001 PAU89 U A/nus cordata Italy 2000 PAU142 V Alnus glutinosa Slovenia 2000 PAU188 U Alnus: incana BeLgillnl 2001 PAUb3 U" Alnus glu/in osa NE France 2005 PAVe&' U A/nus glutinosa NE France 2005 PAUlS7 LI A/nus glutinosa Betgiulll 2001 PAU96 U Alnus g/utinosa Hungary 1999 PAU97 U A/nus glutinosa sail Hungary 1999 PAU98 U A/nus g/utinosa soil Hungary 1999 l'AUb3 U' Alnus glutinosa NE France 2005 Phytophthora cambivora PQC17 C2 Quercus sp. soil NE France 1999 PC643 Cl C.astanea sativa soil SWFranee 2000 PC463 C2 Castanea sativa SW France 1994 PCgal C2 Quercus sp. soil NE France 1999

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd

178 Développement de microsatellites chez Phytophthora alni spp. alni

PRIMER NOTE 135

Table 1 Contirwed

Geographical Year Species lsolate Mitotype Host origin isolated

l'C99428 C2 Castanea satim France 1999 l'CsSr1 C2 Quercus petraRa SWFranCè 1999 l'C627 C2 Castl/nea SIItina ltaly 1999 l'Cla21 C2 Quercus sp. soil NWFrance 1999 l'C4n444 C2 Castanea satina NVvFrance 2004 PC4n1125 C2 Castanea saUva NWFrance 2004 l'COS 1422 C2 Castanea sati:va NVV France 2005 PhytophtllOra{ragariae var. fragariae l'FF309 FF Fragaria x artal1l/SSa Great Britain 1962 l'FFCSL FI' Fragaria xanllnl/Ssa Unknown IJnknown l'FF209 FF Fragaria x artanllSsa England 1946 Phytophthora fragariaR var. ru!>i l'FR109 FR Ru!>u$ sp. Great Britain 1991 PFR163 FR Rubus sp. France Unknown PFR59 FR Rubus sp. Scotland Unknown l'FRCSL FR Rubus sp. England Unknovvn l'FR96795 FR Rubus sp. Scotland 1985 Phytophthora rnctorum CAC4810 ad Unknown France Unknown Phytophthora dnnamomi DSF970060 nd Quercus suber France 1999 Phytophthora d meola AU[045 nd Alnus glutinosa France 1999 Phytophthora dtrophthora 2N1021 nd Rosa sp. France 2002 Phytophthora cryptogea 990675 nd Actinidia sinensis France 1999 Phytophthora erythroseptica 960713 nd Polygonum oberti France 1999 Phytophthom europaea ALS nd Quercus sp. soil France 1998 Phytophthora gonapodyides Gonap4 nd Quercus sp. soil France 1998 Phytophthora humieola 3N1245j nd Ah/us glutirLOsa soi! France 2003 Phytophthora inundata 9500802 nd Alnus glutinosa soil France 1998 Phytophthora Iateralis 98093-1 l'V nd Chamaecyparis sp, France 1998 Phytophthora megasperma BK1 nd Quercus sp. soil France 1998 Phytophthora nicotianae 960.579 nd NicotiallŒ tabacwll France 1996 Phytophthora taxon forest soi! 8CARPl'OC1 nd Quercus sp. soil France 1998 Phytophthora l}{ûmwora 970423 ud Hederasp. France 1997 Phytophthora parasitica 970029 ud Lycopetsi CiY11 sp. France 1997 PhytophtilOra taxon Pgchlamydo Haye,3,1 nd Quercus sp. soil France 1998 Phytophthorapseud0$.'fYingae EW.5 nd Quercussp. soil France 1998 Phytophthom psychrophilll FF20 nd Quercus sp. soil France 1998 Phytophthora qtlercina FNA ad Quercus sp. soH Frt'lncp. 1999 Phytophthora ramorum 2N083 nd Rhododendron sp. France 2002 Phytophthora sojae 443 nd Glycîne1l1ax Unknown Unknown Phytophthorasyringae 2)Z2 nd Quercus'Sp. soil France 1999

" isolateused for microsatellite isalatilm; nd, not determined.

template DNA, 2 pL of lOx polymerase buffer, 1.5 111.1....1 unclear patternandwere therefore discarded. The remaining MgCI;y 0,2 mM dNTP, 0.2 ~lM of each prim.er, 0.5 U ofTaq 10 primer pairs were tested by PŒon a larger collection of polymerase (Sigma) and water wasadded t020 ltL. Touch­ Paa, Pau, Parrt, P. cambiFora and P. fragariae isolates as weil down PCR conditions were fin;t {ive cycles at 94 oC for as \vith 23 other Phytophthora spp. (Table 1). 30 s, 66-62 oC (see initialanneaUng temperatures, Table2) Except PAJ2-FIR that were specifie to Paa and Pam but for 30 sand72 Oc for 1 min; followed byScycles at 94 "C for gave no size polymorphism, all the primers retained yielded 30 s, 66-55 "C minus 0.5 oC per cycle for 30 s, ànd 72 oC for an ampUconwith Paa, Pau, Pam, Pc and PI, confirmingtheir 1 min; 22 cycles at 9,t oC for 30 s, 62-51 OC for 30 s, and 72 oC close kinship. ln Paa, the numberof alleles for the nine for 1 min; and a final elongation sœp at 65 oC for 30 min. AlI polymorphie loci ranged from two to five indicating a low retained forward primers were labelled tQ alI()w size ànd level of alleUç variation, consistent with this hybrid species dye multiplexing. Allek sizes were determined on a CEQ being of recent origin (Brasier et al. 2004; 1000et al. 2006). In SOOOGenetie AnaJysisSystem (Beckman Coulœr). Eightout addition, all the alleles observed in Paa were those encoun­ of the 18 primer pairs tested yielded no polymorphism or tered either in Pam or in Pau (l'able 2), strengthening the

© 2006 The Authors Journal coEilpilation©2006 Blackwell Publishing Ltd

179 ..... ~ 00 o ~ ..... w ~ (1;) 'll'" N :;;:1 i: tT:I :;;:1 Z o..., tT:I Table 2 Primer sequences and characteristics for the nille polymorphie tnicrosatellite loci isolated from Phytophthora alni ssp. aini. Ali reverse l'CR primers Were fluoresœntly labelled

Total numberof different allèles/no. of alleles per individual (allele sizes)

Repeat GenBank Size Annealing P. alni ssp. P. alnIssp. P. fragm'Iae P. fragariae other Locus motif Accession no. Primer sequence (5'-3') O>p) TOC' P. altIi ssp, altIi muHiformis unifomtis P. cam biv9ra var frdgariae var rubi species

PM (CAI\,)Gl~ DQ665899 F: Cl'n::;GA:1'AGAGCCG'I'Cor'rC 201 62 qc 3/2+ (198, 2/1-2+ 2/1-2 4/3 (2)0,211, 1/1 (217) 1/1 (217) (CAA)s R: 'l'.cCrrACllGfITGGGAGCAA.GG 212,216) (198,216) (198,212) 213,215) pA6 (Cl\A)n DQ6659()0 F:AACACCGCG'I'TGAA~CG 303 66 oC 3/3t (278, 1/1 (27$) 3/3 (278, 2/2 (278,281) 2/1-2 <275, 2/2 1'AA(CAA) R: G'l'AGCCACÇ(;c»;CA1'GAPlI'C 284,287) 281,287) 288) (275,288) pA8 (ACA)5 DQ665901 F: GGrCAGCCAA~GCAAJI.GAG 376 66°C 5/3-5+ (356, 3/2-3+ (356, 2/2 2/1-2 (367, 2/2 (364, 1/1 (367) P.europaea 1'(ACà) R: C1'GTC.J\GCTC',Q'\..,)\(;AAGCAG 359,361, 359,3(8) (361,367) 369) 367) (368) 367,368) l'An (CAA)FAG DQ665902 F: 1'GCAAACAG'l'GCG1'crC'1'1'C 226 62°C 2/2+ 2/1-2+ 1/1 (226) 2/1-2 (224, 1/1 (226) 1/1 (225) (C'A'i.) R: CC'rAGNl'CCACl:'CN'...AGC'l'..' (226,232) (226,232) 225) l'AH «'l'M···) DQ665904 F: '1'Gr.'F-Af'IACACACACGPAG1'C 'l73 62 oC 3/3+ (173, 2/2+ (173, 1/1 (177) 4/2 (174, 177, 2/2 (168, 2/2 (TI'1'C)4 R: ~Cl3èAGl'C."1'œ.GA~G 175,177) 175) 179,1811 174) (175.177) pA17 (GTC)4( ... ) DQ665905 F: AGC~CM1'GCAGGAAGC 317 62 oC 2/2 (313,317) 2/2 (313, 1/1 (313) 2/1-2 (309, 2/2 (308, 1/1 (315) P. europtUa '-< C1'G1'C'IGGGCA1'TC~.TGl'C'G 317) 313) 312) (385) o (GC)4 R: c PA23 (GA.")qG~ DQ665906 F: G~~TAGCG\,CG:i'.~CACC 155 62.GCATCQ:;:rcTAAAc.'GAC 148) n o pA30 (AC)8 DQ665907 F:TAGGGGACTI'AMC('CCA 127 66°C 4/4+ (120, 3/3t (120, 1/1 (123) 2/2 (118, 122) 2/2 (120, 2/2 3 'll R: crrTGGCGGA01JA~ ...GA!:r-I'T 122, 123, 124) 122,124) 122) (120,122) ~ pA3I (CAMa DQ665908 F: GCTI'C'TCÀCTGCAÇAGCJl..".C 195 62 oC 2/1-21 1/1 (192) 2/2083. 4/3-4 (lï4, 1/1 (150) 1/1 (150) o' (183,192) 192) 180,186,192) :l R: AGG.Gl'ATTGGAXCTGATGC 61 g "initial annealing temperature for touchdown l'CR; +, loci with alldes consistently showing differentpeak areas. 0'> ~ ~ ; 61 =0(D' l'..J ;yS: o--~ ~g 8':> .~ ~ L"'8 li rh Développement de microsatellites chez Phytophthora a/ni spp. alni

PRIMER NOTE 137

hypothesisfhat Pau and Pam mayactually be the progenitors & its variants: designation of emerging heteroploid hybrid ofPaa. Duetotheunœrtain nature al' ploidy in these different pathogens spreading on Alnus trees. Mycological Research, lOS, taxa, expected heterozygosities and linkage disequilibria 1172-1184. were not computed. Hawever, depending on the locus, Christiansen DG (2006) A microsatellite-based method for genotyping diploid and triploid water frog oftheRal1a escu/ento up ta three Or five different alleles were simultaneausly hybrid complex. Molecular Ecology Notes, 5,190-193. observed in iItdividual Pam and Paa isolates, respective!y. Dutech Ci Amselli'.lll L, Billotte N,Jarne P (2000) Characterization In addition, genatypingofthe PCR products obtainedwith of (GA). l11icrosatellite loci using an enrichment protocol in the the primer pairs PA3, PA6, PAS, PAll, PA14, PA23, PA30 neotropical tree speciesVouacapouaamenCill1a. Molecular Ec%glj. '1lld PA31 shawedthat for individual Paa and Pam isolates, 9,1433-1435. the respective quantifies of the different andes (based on EstoupA, Solignac M, Harry M, CornuetJM (993) Characteriza·· tion of(GT). and (CT), rnicrosate1lites intwo insect species: Apis peak areas) were notalways equal. The occurrence of melliferll and Bombus terresms. Nuclcie Adds ReSelil-cil, 21, 1427­ unbalanced peak ratios suggests these species are paly­ 1431. plaid (Christiansen 2006). The genotype amplification Ioos R, Andrieux A, Marçais B, Frey P (2006) Genetic characteriza­ patterns for these loci will be particularly useful to unravel tion of the natural hybrid species Phytophthora a/rd as inferree! the genotypes of Paa and Pam. fromnuc!ear and mitochondrial DNA analyses. FU11gal Gelleties tmd Biol6gy, 43, 511-529. Rozen S, Skaletski H (2000) PRIMER3 On the 'vVV\i1;V for general References userB and for biologist programmers. In: BioinjiJntllltics A1ethods Brasier CM, Cooke DEL, DuncanJM (1999) Origin ofa uew Phyto­ Ilnd Prot6CO!S: Methods i11 Molecular Biology (eds Krawetz S, phthora pathogenthroughinterspecific hybridization. Proceedings Mi,ener S), pp. 365-386. Hurnana Press, Totowa, NewJersey. ofthe National Academy ofSciences, USA, 96, 5878-5883. Weising K, Nybom H, Wolff K, Meyer W (1995) DNA Fil1gerprinl.tllg BrasierCM,KirkSA,De1canJet al. (2004) Phytophtho1"a aini sp. nov. in Plllnts and FU11gi. CRC press, Boca Raton, Florida.

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd

181 Monsieur BARRES Benoît

DOCTORAT de l'UNIVERSITE HENRI POINCARE, NANCY 1 en BIOLOGIE VEGETALE & FORE5TIERE

Vu, APPROUVÉ ET PERMIS D'IMPRIMER ~~,jlJ

Nancy, le .?~Pb

Le Président de l'Université

03f0œ~œf0œ~ Université Henri Poincaré, Nancy-I 24-30 rue Lionnois· B.P. 3069 - 54013 NANCY CEDEX Tél. : 03 83 85 48 00 - Fax: 03 83 85 48 48