Vous êtes sur la page 1sur 371

Universit Paris-Sud XI U.F.R.

Scientifique dOrsay

Thse
Prsente pour lobtention du grade de

Docteur en Sciences de lUniversit Paris-Sud XI


Spcialit : Sciences du Vgtal

par Jan Drouaud

Etude de la distribution des vnements de recombinaison miotique chez Arabidopsis thaliana

Soutenue le 29 juin 2010 devant la commission dexamen : Dominique de Vienne Prsident Anne-Marie Chvre Rapporteur Bernard de Massy Rapporteur Denise Zickler Examinateur Thierry Langin Examinateur Christine Mzard Directrice de Thse

Remerciements
Je tiens remercier en premier lieu ma directrice de thse, Christine Mzard, pour mavoir accueilli dans son quipe et propos de relever des dfis scientifiques et techniques palpitants, mais aussi pour ses qualits humaines remarquables. Je souhaite galement exprimer toute ma gratitude envers les membres du jury dvaluation de mon travail de thse : Dominique de Vienne, Anne-Marie Chvre, Bernard de Massy, Denise Zickler et Thierry Langin. Je voudrais ensuite tmoigner ma reconnaissance Philippe Guerche et Georges Pelletier, qui mont soutenu sans conditions depuis de nombreuses annes dans mon effort damlioration de mes comptences et connaissances. Je remercie bien sur mes collgues dquipe, Laurne Giraut et Hossein Khademian, pour leur contribution trs efficace lavancement de nos projets de recherche. Naturellement, je remercie chaleureusement Raphal Mercier, Mathilde Grelon, Fabien Nogu, Eric Jenczewski, Christine Horlow et tous les autres membres du groupe Miose et Recombinaison (Anouchka, Arnaud, Aurlie, Ccile, Daniel, Florence, Ghislaine, Laurence, Laurie, Marta, Nathalie, Nicolas, Nicole, Pierre, Sylvia, Sylvie, Wayne), pour les nombreuses conversations trs sympathiques et enrichissantes qui font de notre petite communaut un monde un peu part. Je suis plus gnralement redevable tous les habitants de la SGAP, pour leur gentillesse, leur disponibilit, et leur comptence, toutes qualits qui ont grandement facilit ma tche.

Abrviations
ADN CDB CO CS dJH DL HDBS : : : : : : : acide dsoxyribonuclique. cassure(s) double brin. crossover(s). complexe synaptonmal. jonction de Holliday double. dsquilibre de liaison. hybridation de brin dpendante de la synthse. insertion ou dltion. non-crossover(s). nodule(s) de recomninaison prcoce(s). nodule(s) de recombinaison tardif(s). point(s) chaud(s). point(s) chaud(s) historique(s). raction polymrasique en chane. rparation des CDB. jonction de Holliday simple. polymorphisme de substitution nuclotidique.

INDEL : NCO NRP NRT PC PCH PCR RCDB sJH SNP : : : : : : : : :

Liste des tables et figures


Table 1. Relation entre longueur du complexe synaptonmal et nombre de crossings-over. ........ 29 Table 2. Relation entre longueur des axes et nombre de nodules de recombinaison prcoces. .... 30 Table 3. Liste des points chauds de CO tudis par typage de sperme chez les mammifres. ...... 46 Table 4. Liste des points chauds de CO et NCO tudis par typage de sperme chez les mammifres. ............................................................................................................................................... 48 Table 5. Distributions du nombre de crossovers par chromatide et du nombre de chiasmas par bivalent dans les mioses mle et femelle chez Arabidopsis thaliana. ................................................... 64 Table 6. Distributions observe et attendue du nombre de chiasmas non-obligatoires par bivalent dans les mioses mle et femelle chez Arabidopsis thaliana. .................................................................. 64 Figure 1. Cycle de dveloppement des eucaryotes.................................................................................. 8 Figure 2. Chiasmas....................................................................................................................................... 9 Figure 3. Sgrgation chromosomique en mitose et en miose. ......................................................... 10 Figure 4. Formation des lments axiaux. .............................................................................................. 11 Figure 5. Bouquets. .................................................................................................................................... 11 Figure 6. Structure du complexe synaptonmal (I) ............................................................................... 12 Figure 7. Structure du complexe synaptonmal (II). ............................................................................ 12 Figure 8. Nodules de recombinaison prcoces et initiation du synapsis. .......................................... 13 Figure 9. Localisation de la protine RPA.............................................................................................. 14 Figure 10. Evolution de la composition des nodules de recombinaison. .......................................... 15 Figure 11. Nodules de recombinaison tardifs........................................................................................ 17 Figure 12. Chiasma et foci MLH1. .......................................................................................................... 17 Figure 13. Modle de la double jonction de Holliday, de Szostak et al .............................................. 18 Figure 14. De la formation de la cassure double-brin linvasion de brin........................................ 20 Figure 15. Voies de recombinaison homologue miotique ................................................................. 24 Figure 16. Le nombre de chiasmas par miose nest pas alatoire. .................................................... 27 Figure 17.Relation entre taille physique des bivalents et nombre de crossings-over. ...................... 28 Figure 18. Fusion chromosomique et interfrence des crossings-over chez C. elegans. ................... 35 Figure 19. Foci MLH1 au stade pachytne. ........................................................................................... 36 Figure 20. Relation entre les distributions de nombre et de position des crossings-over sur les bivalents....................................................................................................................................................... 37 Figure 21. Le biais de transmission alllique du la conversion gnique.......................................... 40

Figure 22. Distribution des CO dans la rgion a1-yz1 chez le mas.................................................... 49

Sommaire
1 Introduction......................................................................................................................................... 8 1.1 1.2 1.2.1 Prambule ................................................................................................................................... 8 Cytologie et biologie molculaire de la miose...................................................................... 9 Description cytologique de la premire division de miose............................................ 9 La prophase................................................................................................................. 10

1.2.1.1

1.2.1.1.1 La formation des axes chromosomiques........................................................... 10 1.2.1.1.2 Lalignement prsynaptique des homologues................................................... 11 1.2.1.1.3 La mise en place du complexe synaptonmal................................................... 12 1.2.1.1.3.1 Les sites dinitiation du synapsis ................................................................. 12 1.2.1.1.3.2 La relation entre linitiation du synapsis et linitiation de la recombinaison homologue ................................................................................................ 13 1.2.1.1.3.3 Les nodules de recombinaison prcoces ................................................... 14 1.2.1.1.3.4 Les constituants du complexe synaptonmal............................................ 14 1.2.1.1.3.4.1 Les protines des lments latraux ................................................... 14 1.2.1.1.3.4.2 Les protines de la zone centrale et les facteurs dinitiation du synapsis 15

1.2.1.1.3.5 Les entremlements de bivalents................................................................ 16 1.2.1.1.4 Pachytne ............................................................................................................... 16 1.2.1.1.5 Diplotne et diacinse.......................................................................................... 17 1.2.1.2 1.2.2 Mtaphase, anaphase et tlophase ........................................................................... 18

Description molculaire de la recombinaison miotique .............................................. 18 Formation des cassures double-brin ....................................................................... 19 Maturation des CDB.................................................................................................. 19 Invasion de brin.......................................................................................................... 20 Le biais inter-homologues......................................................................................... 21 Les voies de recombinaison homologue................................................................. 23

1.2.2.1 1.2.2.2 1.2.2.3 1.2.2.4 1.2.2.5

1.2.2.5.1 Les crossovers de type I ...................................................................................... 24 1.2.2.5.2 Les crossovers de type II..................................................................................... 25 1.2.2.5.3 Les non-crossovers............................................................................................... 26 1.2.2.5.4 Les intermdiaires de recombinaison aberrants ............................................... 26 1.2.2.6 1.3 1.3.1 La rsolution des intermdiaires de recombinaison.............................................. 26

Le nombre et la distribution des vnements de recombinaison miotique................... 27 Le nombre de crossings-over ............................................................................................ 27 La dtermination du nombre de crossings-over.................................................... 27 La dtermination du nombre dintermdiaires de recombinaison ...................... 29

1.3.1.1 1.3.1.2 1.3.2

Linterfrence des crossings-over...................................................................................... 29 Linterfrence gntique............................................................................................ 29 Linterfrence cytologique......................................................................................... 32 Les variations de lintensit de linterfrence ......................................................... 34

1.3.2.1 1.3.2.2 1.3.2.3

1.3.2.3.1 Y a-t-il une mtrique descriptive de linterfrence ? ........................................ 34 1.3.2.3.2 Les modles biologiques de linterfrence ........................................................ 35 1.3.3 La distribution des crossings-over .................................................................................... 36 Le dterminisme chromosomique de la localisation des crossings-over ........... 36 Le dterminisme gnomique de la distribution des crossings-over .................... 38

1.3.3.1 1.3.3.2 1.3.4 1.4 1.4.1 1.4.2

La distribution des cassures double-brin ......................................................................... 40 Les points chauds de recombinaison miotique ................................................................. 41 Les points chauds de recombinaison miotique chez les champignons ..................... 41 Les points chauds de recombinaison miotique chez les animaux et les plantes ...... 44 Lidentification des premiers points chauds de mammifres............................... 44 Les points chauds de mammifres caractriss par typage de sperme ............... 45 Les points chauds de plantes .................................................................................... 47 Le projet HapMap et les points chauds historiques chez lhomme.................... 49

1.4.2.1 1.4.2.2 1.4.2.3 1.4.2.4

1.4.2.5

Vers llucidation du dterminisme de localisation des points chauds de

mammifres................................................................................................................................... 50 2 Rsultats ............................................................................................................................................. 53 2.1 2.2 Les voies des crossovers : la parole est aux plantes ............................................................ 53 Distributions de crossovers spcifiques du sexe et variations du niveau de linterfrence

le long du chromosome 4 dArabidopsis thaliana ................................................................................ 63 2.2.1 2.2.2 2.2.3 2.3 Synthse de la publication.................................................................................................. 63 Rsultats additionnels ......................................................................................................... 64 Publication............................................................................................................................ 64 Diversit des profils de crossover et de conversion gnique des points chauds

miotiques chez Arabidopsis thaliana. ................................................................................................... 77 2.3.1 2.3.2 3 Synthse de la publication.................................................................................................. 77 Publication............................................................................................................................ 78

Discussion et perspectives.............................................................................................................105 3.1 3.2 3.3 Htrochiasmie et variations de longueur du complexe synaptonmal.........................105 Interfrence et distribution des crossovers le long du chromosome 4..........................106 Variations de la concidence des crossovers et intensit de linterfrence le long du

chromosome 4 .....................................................................................................................................106 3.4 3.5 3.6 4 Distribution atypique des crossovers au point chaud 130x.............................................108 Crossovers et non-crossovers au point chaud 130x dans le mutant Atmsh4 ................109 Bilan .........................................................................................................................................110

Matriels et mthodes.....................................................................................................................112 4.1 Characterization of meiotic crossovers in pollen from Arabidopsis thaliana...................112

5 6

Rfrences bibliographiques..........................................................................................................146 Annexe..............................................................................................................................................176 6.1 La variation des taux de crossing-over le long du chromosome 4 dArabidopsis thaliana

rvle la prsence de points chauds de recombinaison miotique .........................................176

Figure 1. Cycle de dveloppement des eucaryotes. Les eucaryotes se reproduisant par voie sexue ont un cycle de dveloppement caractris par une alternance de phases haplode et diplode. Des cellules diplodes diffrenties accomplissent la miose en subissant deux sgrgations chromosomiques lissue dune seule phase de rplication de lADN. Les cellules haplodes rsultantes ne contiennent quun exemplaire de chaque chromosome. Elles se diffrentient en gamtes directement, ou aprs une phase de multiplication par division mitotique. Deux gamtes de types sexuels diffrents (rouge ou bleu) fusionnent pour former le zygote diplode, qui peut entamer un srie de divisions mitotiques, ou bien entreprendre la miose directement. (Kohli et Hartsuiker 2008).

1 Introduction
1.1 Prambule
Miose et fcondation sont le Yin et le Yang du cycle de dveloppement des organismes eucaryotes. La fcondation, ou syngamie, consiste en la fusion de deux cellules dites gamtiques. Chez les plantes et les animaux, ces cellules sont diffrenties morphologiquement en deux types parentaux, mle et femelle. Mais chez de nombreux protistes et champignons cette anisogamie nexiste pas et le nombre de types sexuels peut dpasser deux. Dune manire gnrale, les cellules ralisant la syngamie possdent un seul exemplaire de chaque chromosome : elles sont haplodes 1 . Lors de la fcondation, les noyaux des deux cellules parentales fusionnent, de sorte que les cellules qui en sont issues sont diplodes. La miose est un processus symtrique de la syngamie en ce quelle rduit la plodie : elle restore lhaplodie (figure 1). Elle accomplit pour cela la sgrgation des exemplaires homologues de chaque chromosome dans des cellules spares. Cette sgrgation rsulte donc dune division cellulaire trs spcialise, la premire division miotique, que lon qualifie de rductionnelle. A linstar dune mitose, cette division est prcde dune phase de synthse dADN correspondant la duplication de chaque chromosome en deux chromatides surs. Ces chromatides surs ne sont spares que lors de la deuxime division miotique, qualifie dquationnelle (figure 3). Dun point de vue gntique, la sgrgation alllique nest complte quaprs la division quationnelle. Cela est du au phnomne de crossing-over, qui est non seulement lun des fondements de la recombinaison gntique, mais galement la condition sine qua none de la sgrgation rgulire des homologues lanaphase de la premire division miotique, dans la plupart des organismes eucaryotes 2 . Le crossing-over est un processus de recombinaison molculaire homologue, qui aboutit lchange rciproque de segments entre deux chromatides. Il se produit dune manire programme au cours de la prophase de la premire division miotique et implique dans la plupart des cas des chromatides homologues non surs . La manifestation cytologique la plus vidente du crossing-over est le chiasma. Ces figures dentrecroisement daxes chromosomiques

1 2

Le cas des individus et des espces polyplodes ne sera pas abord dans ce manuscrit. Les cas de sgrgation achiasmatique, chez la drosophile par exemple, et de disjonction distributive ne seront pas traits.

Figure 2. Chiasmas. Bivalents longs mtacentriques (L) et mdians tlocentriques (M) chez Omocestus viridulus au stade diacinse. A droite, dans la srie de reprsentations schmatiques, les centromres sont figurs par des ronds noirs et les chromatides par des traits noirs. (Southern 1967).

homologues sont observables aux stades diacinse et mtaphase de la premire division miotique (figure 2). Le terme de crossover (CO) est utilis dans ce manuscrit pour dsigner la consquence molculaire dun crossing-over. Les CO ne sont donc en principe observable que dans le produit de la miose dun hybride, ou bien dans sa descendance. Lobjectif de cette section dintroduction est de prsenter le contexte, c'est--dire les structures et phnomnes cellulaires et molculaires pertinents, ainsi que les manifestations de la recombinaison homologue miotique. La structuration de cette partie fait donc apparatre des choix dlibrs : Les aspects de la miose nayant a priori pas de rapport avec la recombinaison seront peu ou pas abords. Les causes et les consquences immdiates de la recombinaison miotique seront considres principalement, et seules quelques questions dordre volutif seront voques. Les divers eucaryotes ont invent une grande varit de mcanismes biologiques permettant de former au moins un chiasma par paire dhomologues. Les processus communs la majorit des organismes seront dcrits, mais les particularits de chaque espce ne seront pas toujours mentionnes.

1.2 Cytologie et biologie molculaire de la miose


Le modle actuel de recombinaison homologue des chromosomes au cours de la miose a t labor progressivement depuis la redcouverte des lois de Mendel en 1900. La combinaison dapproches gntiques et cytologiques a tout dabord permis, dans les premires dcennies du 20 sicle, dlaborer la thorie chromosomique de lhrdit (Sutton 1902; Morgan 1911; Darlington 1931). Puis, la dcouverte de la structure de lADN et de son mode de rplication, ainsi que le dveloppement de la gntique des champignons, ont fourni les moyens dlaborer des modles molculaires de la recombinaison homologue miotique.

1.2.1 Description cytologique de la premire division de miose


La premire division miotique se distingue dune division mitotique plusieurs niveaux : La prophase est trs longue (22 33 heures chez Arabidopsis thaliana (Armstrong, Franklin et al. 2003; Stronghill et Hasenkampf 2007; Jones et Franklin 2008), plus de 24 heures chez le mas

Figure 3. Sgrgation chromosomique en mitose et en miose. Dans des cellules mitotiques, la cohsion entre chromatides surs disparat lanaphase, ce qui permet leur sgrgation (i-iii). Quand les kintochores sattachent aux microtubules manant des ples opposs, les forces dattraction exerces par le fuseau sont contrebalances par la cohsion entre les centromres des chromatides soeurs (iii). La tension rsultante stabilise lattachement des microtubules. Quand toutes les paires de chromatides sont bi-orientes, la cohsion disparat et les chromatides sgrgent aux ples opposs (iv). Durant la prophase de premire division de miose, les homologues sapparient et forment des chiasmas (v-vii). La combinaison des chiasmas et de la cohsion permet dorienter les homologues sur le fuseau, de la mme manire que les chromatides surs salignent pendant la mitose (vii). Quand les cohsines sont dgrades, les homologues sgrgent (viii). Une nouvelle phase de sgrgation se produit ensuite sans rplication pralable (ix-x). La cohsion entre les centromres est maintenue jusquen anaphase de deuxime division. Les flches bleues indiquent les forces de traction exerces par les microtubules. Les lignes pointilles correspondent aux plans de division cellulaire. (van Veen et Hawley 2003; McDougall, Elliott et al. 2005).

(Anderson, Stack et al. 1988)). A titre de comparaison, la prophase de mitose dure 55 min dans des cellules de pointe racinaire doseille (Kuroki et Tanaka 1973), moins de 40 minutes dans des cellules racinaires de diverses espces de Crepis (Langridge, O'Malley et al. 1970). De plus, la prophase peut tre subdivise en une srie de stades bien reconnaissables, dfinis par des transitions de la morphologie et/ou de lorganisation spatiale des chromosomes. Lors de la mtaphase, les chromosomes homologues sont associs via la combinaison des chiasmas et de la cohsion entre chromatides surs. Les kintochores des chromatides surs sont mono-orients, cest dire quils sont tous deux relis un mme ple du fuseau, tandis que les kintochores des homologues sont relis aux ples opposs. A lanaphase, la cohsion entre chromatides surs disparat le long des bras mais persiste au niveau des centromres, ce qui aboutit la sparation des homologues puis leur migration dans des directions opposes (Page et Hawley 2003; Hochwagen 2008)(figure 3).

1.2.1.1 La prophase
Au cours de la prophase de miose, les chromosomes homologues sindividualisent puis sassocient troitement sur toute leur longueur. La structure caractristique qui rsulte de ce processus de synapsis est appele le complexe synaptonmal (CS). Les chromosomes homologues ainsi apparis sont alors dnomms bivalents. Le CS est observ chez tous les eucaryotes dont la miose a t tudie, lexception de quelques champignons comme Schizosaccharomyces pombe et Aspergillus nidulans (Page et Hawley 2004). Avant la mtaphase, le CS se dissocie. Les axes des homologues se sparent et les chiasmas se rvlent alors comme les seules connexions subsistant entre eux.

1.2.1.1.1 La formation des axes chromosomiques


Au stade leptotne la chromatine se condense. Les chromosomes sindividualisent sous la forme de fibres fines (du grec leptos mince ), observables en microscopie photonique. La microscopie lectronique rvle des segments initialement courts et nombreux, qui sallongent et se connectent (Albini et Jones 1987; Dresser et Giroux 1988)(figure 4). Ces filaments sont constitus dun lment axial le long duquel la chromatine sorganise en boucles (Rattner, Goldsmith et al. 1981). Les axes chromosomiques sont galement visualisables par immunomarquage, en utilisant des anticorps dirigs contre des protines constitutives des axes, comme HOP1 chez Saccharomyces cerevisiae, ASY1 chez les plantes (Armstrong, Caryl et al. 2002;

10

Figure 4. Formation des lments axiaux. Sur ce clich dun talement de noyau dAllium cepa au stade leptotne, color au nitrate dargent, on observe de nombreux fragments daxes non apparis. La barre reprsente 10 m. (Albini et Jones 1987).

Figure 5. Bouquets. A Coupe optique dun noyau dovocyte de vache au stade leptotne/zygotne. Le plan focal est lquateur du noyau. Les tlomres sont marqus en rouge, les lments axiaux (protine SCP3) en vert, et lADN en bleu. La barre reprsente 5 m. (Pfeifer, Scherthan et al. 2003). B Microphotographie en contraste de phase dun noyau dOedipina uniformis au dbut du stade zygotne. La barre reprsente 10 m. (Kezer, Sessions et al. 1989).

Mikhailova, Phillips et al. 2006), ou encore les cohsines (Prieto, Tease et al. 2004; Chelysheva, Diallo et al. 2005). Cest galement au cours du stade leptotne que la recombinaison miotique est initie. Ainsi, la protine RAD51, qui est implique dans les tapes prcoces de la recombinaison homologue miotique (section 1.2.2.2), est localise ds avant le dbut du synapsis sous forme de nombreux foyers distribus le long des axes, aussi bien chez lhomme (Brown, Judis et al. 2005), la souris (Barlow, Benson et al. 1997), le lys (Terasawa, Shinohara et al. 1995) et A. thaliana (Mercier, Armstrong et al. 2003).

1.2.1.1.2 Lalignement prsynaptique des homologues


Dans certaines espces un alignement prsynaptique des chromosomes homologues est observable vers la fin du leptotne. Ainsi chez Sordaria macrospora les axes sont trs clairement parallles, distants denviron 300 nm (Tesse, Storlazzi et al. 2003). Dans la majorit des espces par contre, il existe peu ou pas dvidence cytologique dun rapprochement prcoce des homologues sur toute leur longueur ds le stade leptotne. Nanmoins, on peut trouver des preuves indirectes dun alignement prsynaptique des homologues, par exemple chez un Allium triplode (Loidl et Jones 1986), qui laissent penser que ce phnomne pourrait tre en fait discret mais rpandu, ventuellement universel. Dans la plupart des espces, une voire les deux extrmits tlomriques des axes sont ancres lenveloppe nuclaire la fin du stade leptotne. La transition vers le stade zygotne est caractrise par le regroupement, plus ou moins marqu selon les cas, des tlomres, de sorte que les axes convergents forment une figure voquant un bouquet (Zickler et Kleckner 1998; Scherthan 2001)(figure 5). Chez A. thaliana cet arrangement caractristique nest pas retrouv : les tlomres semblent plutt associes au nuclole (Armstrong, Franklin et al. 2001). Le regroupement des tlomres perdure pendant le stade zygotne, puis disparat au cours du pachytne (Zickler et Kleckner 1998). Outre le bouquet, un couplage des centromres a t observ chez S. cerevisiae, qui semble impliqu la fois dans linitiation du synapsis (Tsubouchi et Roeder 2005), et dans la sgrgation des homologues en labsence de chiasmas (Kemp, Boumil et al. 2004). Il est trs vraisemblable que le bouquet, sous ses formes canonique ou htrodoxe, et ventuellement dautres phnomnes de mouvement des chromosomes, comme le couplage des centromres chez S. cerevisiae, jouent un rle dterminant dans linitiation du synapsis. Le rapprochement des chromosomes favoriserait les interactions initiales entre homologues. Ces

11

Figure 6. Structure du complexe synaptonmal (I). Sur cette reprsentation en coupe transversale du complexe synaptonmal sont figurs: Les lments latraux, contenant un arrangement hypothtique de protines (bleues et vertes). La rgion centrale, comprenant les filaments transversaux constitus de dimres de protines (jaunes et oranges), et llment central. La chromatine, organise en boucles manant des lments latraux. (Page et Hawley 2004).

Figure 7. Structure du complexe synaptonmal (II). Sur ce clich dune portion de CS de Blaps cribosa en vue frontale, color lactate duranyle, on distingue les lments latraux (LE), llment central (CE), les filaments transversaux (TF) et la chromatine (ch). (Schmekel et Daneholt 1998).

contacts entraneraient lalignement des rgions proches, de sorte que les axes se juxtaposeraient progressivement, dune manire synergique (Schlecht, Lichten et al. 2004). Par ailleurs, il existe des preuves convaincantes dun couplage entre les mcanismes dalignement prsynaptique des homologues et dinitiation de la recombinaison homologue au moins chez S. cerevisiae (Weiner et Kleckner 1994; Peoples, Dean et al. 2002) et S. macrospora (Storlazzi, Tesse et al. 2003; Storlazzi, Gargano et al. 2010). Vraisemblablement, le processus de recherche dhomologie puis dinvasion de brin (par les extrmits des cassures gnres par la protine SPO11, voir la section 1.2.2.1.) dtermine le rapprochement physique des axes des homologues qui prcde immdiatement le synapsis. En revanche, le processus dinvasion de brin semble, assez logiquement, ne jouer aucun rle dans la formation du bouquet chez S. cerevisiae (TrellesSticken, Loidl et al. 1999). Il faut noter cependant que des mcanismes indpendants de la recombinaison homologue semblent galement contribuer lalignement des axes chez S. cerevisiae (Weiner et Kleckner 1994; Cha, Weiner et al. 2000) sans que lon sache de quelle manire. Il a t propos que des interactions ADN-ADN paranmiques (non entrelaces, c'est--dire dissociables sans rotation mutuelle des brins) puissent entrer en jeu, mais lheure actuelle cela nest toujours pas tabli (Weiner et Kleckner 1994). Par ailleurs, dans certains organismes tels que la drosophile et Caenorhabditis elegans lalignement, puis lappariement des homologues sont totalement indpendants du processus de recombinaison homologue (Dernburg, McDonald et al. 1998; McKim, Green-Marroquin et al. 1998; Alpi, Pasierbek et al. 2003; Couteau, Goodyer et al. 2004; Zickler 2006).

1.2.1.1.3 La mise en place du complexe synaptonmal


Durant le stade zygotne (du grec zygon joug ), des filaments protiques transversaux reliant les axes des homologues sont mis en place. Ce processus est appel synapsis et la structure qui en rsulte complexe synaptonmal (CS). Au sein du CS les axes des homologues sont dnomms lments latraux. Les filaments transversaux dfinissent une zone centrale de largeur constante, au milieu de laquelle se trouve un lment linaire plus dense, llment central (figures 6 et 7).

1.2.1.1.3.1 Les sites dinitiation du synapsis


Le synapsis dbute au niveau de sites de rapprochement troit des homologues puis progresse le long des axes. Au tout dbut du zygotne, les sites primitifs dinitiation du synapsis sont peu nombreux et, dans la plupart des espces, plus frquemment localiss prs des extrmits des chromosomes. Cela est observ chez lhomme (Rasmussen et Holm 1978; Brown, Judis et al.

12

Figure 8. Nodules de recombinaison prcoces et initiation du synapsis. Ce clich reprsente une portion de noyau dAllium cepa au stade zygotne, color par lacide phosphotungstique. Les flches indiquent les nodules (a) associs au complexe synaptonmal, ou (b) prsents aux sites dassociation, ou (c) quidistants dlments axiaux troitement aligns. La barre reprsente 5 m. (Albini et Jones 1987).

2005), le bombyx (Holm et Rasmussen 1980), le criquet (Jones et Croft 1986), S. macrospora (Zickler, Moreau et al. 1992), le bl (Corredor, Lukaszewski et al. 2007), le mas (Burnham, Stout et al. 1972), le seigle (Abirached-Darmency, Zickler et al. 1983; Gillies 1985), deux espces dAllium (Albini et Jones 1987), Crepis capillaris (Banerjee et Jones 1999), Tradescantia (Hasenkampf 1984) et peut-tre A. thaliana (Lopez, Pradillo et al. 2008), mais pas le lys (Rasmussen et Holm 1980) ou S. cerevisiae (Tsubouchi, Macqueen et al. 2008): dans ce dernier cas le synapsis semble plutt dbuter au niveau des centromres. Dans de nombreuses espces vgtales, les sites dinitiation du synapsis sont par la suite observs en beaucoup plus grand nombre des positions interstitielles. Cela semble ne pas tre le cas chez les animaux ou les champignons, y compris dans les espces pourvues de bivalents assez longs (voir rfrences ci-dessus et (Rasmussen et Holm 1980; Zickler et Kleckner 1999)). Lexpansion bi- ou uni-directionnelle (Zickler, Moreau et al. 1992) des zones de synapsis aboutit leur confluence, de sorte que les homologues sont apparis sur toute leur longueur la fin du zygotne (mis part les cas de rarrangements structuraux et dentremlements, voir sections 1.2.1.1.4. et 1.2.1.1.3.5.).

1.2.1.1.3.2 La relation entre linitiation du synapsis et linitiation de la recombinaison homologue


Dans diverses espces vgtales, la microscopie lectronique rvle que des nodules plus ou moins sphriques sont observs le long des axes, mais plus frquemment au niveau des sites de rapprochement des homologues. Parfois des ponts reliant les axes sont galement observs (Albini et Jones 1987; Anderson, Stack et al. 1988; Anderson, Hooker et al. 2001; Stack et Anderson 2002)(figure 8). Plusieurs tudes suggrent par ailleurs que la plupart sinon la totalit de ces nodules, qualifis de prcoces, correspondent aux foyers de dtection immunocytochimique des protines RAD51 et DMC1 chez le lys (Anderson, Offenberg et al. 1997) et la souris (Moens, Chen et al. 1997; Tarsounas, Morita et al. 1999; Moens, Kolas et al. 2002; Moens, Marcon et al. 2007). De plus, les foci RAD51 chez lhumain et la souris sont prfrentiellement observs au niveau ou proximit des sites dinitiation du synapsis (Barlow, Benson et al. 1997). Par ailleurs, plusieurs tudes dmontrent que la protine SPO11, qui permet linitiation du processus de recombinaison via la formation de cassures double-brin (voir la section 1.2.2.1), est galement ncessaire la formation du CS dans la plupart des espces (pas chez Drosophila melanogaster ni C. elegans). Cest le cas par exemple chez S. cerevisiae (Keeney, Giroux et al. 1997), Coprinus cinereus (Celerin, Merino et al. 2000), Mus musculus (Baudat, Manova et al. 2000), A. thaliana (Grelon, Vezon et al. 2001), S. macrospora (Storlazzi, Tesse et al. 2003). Qui plus est les recombinases RAD51 et DMC1, sont galement requises chez S. cerevisiae la formation des associations axiales (Rockmill, Sym et al. 1995). 13

Figure 9. Localisation de la protine RPA. Dans ce spermatocyte humain au stade zygotne, la cohsine SMC3 des axes est marque en vert, les centromres en bleu, et la protine RPA en rouge. La protine RPA est localise dans des foci distribus le long des axes synapss ou non synapss, ainsi que dans des structures formant des ponts entre les axes. La barre reprsente 10 m. (Oliver-Bonet, Campillo et al. 2007).

En rgle gnrale, lassemblage du CS est donc conditionn par linitiation de la recombinaison homologue miotique.

1.2.1.1.3.3 Les nodules de recombinaison prcoces


Tout au long du zygotne, on peut observer des nodules prcoces le long des axes, synapss ou non. Chez les plantes, leur nombre ne varie pas sensiblement entre le dbut et la fin du synapsis (Anderson, Hooker et al. 2001; Stack et Anderson 2002), tandis quil semble diminuer un peu chez la souris (Moens, Marcon et al. 2007). Par contre, leur composition protique se modifie manifestement : RAD51 et DMC1 disparaissent progressivement, tandis quapparaissent les protines RPA (figure 9), ATM, BLM, MSH4 et MSH5 (Plug, Peters et al. 1998; Moens, Kolas et al. 2002; de Boer, Stam et al. 2006; Moens, Marcon et al. 2007; Oliver-Bonet, Campillo et al. 2007)(figure 10). Des distinctions terminologiques ont t introduites afin de distinguer les nodules selon leur ordre dapparition et leur composition (Moens, Marcon et al. 2007) : les nodules prcoces (NRP) apparaissent ds le leptotne et contiennent essentiellement RAD51 et DMC1. les nodules apparaissant au cours du zygotne sont appels nodules de transition ; ils contiennent RPA, ATM, BLM, MSH4 et MSH5 et remplacent progressivement les nodules prcoces. les nodules tardifs (NRT) apparaissent au stade pachytne, en nombre rduit, mesure que les nodules de transition disparaissent, et sont marqus par les protines MLH1 et MLH3 (voir les sections 1.2.1.1.4. et 1.2.2.5.1.).

1.2.1.1.3.4 Les constituants du complexe synaptonmal


1.2.1.1.3.4.1 Les protines des lments latraux Certaines des protines constitutives des lments latraux sont conserves, plus ou moins selon les cas, chez tous les eucaryotes. Cest le cas notamment des protines ncessaire la cohsion des chromatides surs (les cohsines), dont la prsence nest pas strictement requise la formation des axes (Revenkova, Eijpe et al. 2004; Chelysheva, Diallo et al. 2005; Brar, Hochwagen et al. 2009), mais qui dterminent nanmoins leur morphologie et/ou assurent leur intgrit chez S. cerevisiae (Klein, Mahr et al. 1999), la souris (Bannister et Schimenti 2004; Revenkova, Eijpe et al. 2004; Xu, Beasley et al. 2005; Novak, Wang et al. 2008), C. elegans (Pasierbek, Jantsch et al. 2001; Pasierbek, Fodermayr et al. 2003) et A. thaliana (Cai, Dong et al. 2003; Chelysheva, Diallo et al. 2005). Dans tous les eucaryotes galement, des protines domaine HORMA font partie intgrante des lments latraux: Hop1p chez S. cerevisiae (Hollingsworth et Byers 1989), ASY1

14

Figure 10. Evolution de la composition des nodules de recombinaison. Ces graphes indiquent le nombre (et lerreur standard) de foci dans lesquels les protines RAD51 (A), RPA (B), MSH4 (C), MSH5 (D), MLH1 (E) et MLH3 (F) sont dtectes, aux stades leptotne (LEPT), zygotne prcoce (EZ), ou tardif (LZ), pachytne prcoce (EP), intermdiaire (MP), ou tardif (LP). Les comparaisons deux deux des nombres de foci aux diffrents stades sont indiques sous chaque panneau graphique. * p<0,05 ; ** p<0,01 ; *** p<0,001 ; n.s non significatif. (Lenzi, Smith et al. 2005).

chez A. thaliana (Caryl, Armstrong et al. 2000), PAIR2 chez le riz (Nonomura, Nakano et al. 2004), HORMAD1 et HORMAD2 chez la souris (Wojtasz, Daniel et al. 2009), HIM-3 (Zetka, Kawasaki et al. 1999), HTP-1 et HTP-2 chez C. elegans (Couteau et Zetka 2005). Leur absence nentrane gnralement pas de dfaut majeur de constitution des axes. En revanche, dautres protines sont peu conserves, comme Red1p chez S. cerevisiae, qui est faiblement similaire SYCP2 chez M. musculus (Offenberg, Schalk et al. 1998), tandis les protines SYCP3 de souris et HTP-3 de C. elegans (Severson, Ling et al. 2009) ne possdent pas dorthologue dans des eucaryotes lointainement apparents. Ces quatre protines sont nanmoins requises pour la formation des lments latraux, ce qui suggre que leur fonction, dfaut de leur structure primaire, est conserve. Il est possible que des protines homologues sur le plan de la structure tridimensionnelle soient reprsentes dans tous les eucaryotes, qui contribuent toutes former les axes. Toutes les protines des lments latraux, conserves ou non, sont importantes, voire ncessaires au synapsis. 1.2.1.1.3.4.2 Les protines de la zone centrale et les facteurs dinitiation du synapsis Les protines de la zone centrale sont encore moins conserves que celles des lments latraux. Cependant, dans toutes les espces, les filaments transversaux sont constitus de protines analogues sur un plan structural : Zip1 chez S. cerevisiae (Sym, Engebrecht et al. 1993), SYP1, SYP2, SYP3 et SYP4 chez C. elegans (Colaiacovo, MacQueen et al. 2003; Smolikov, Eizinger et al. 2007; Smolikov, Schild-Prufert et al. 2009), SCP1 chez la souris (Meuwissen, Offenberg et al. 1992), ZYP1 chez A. thaliana (Higgins, Sanchez-Moran et al. 2005). Elles possdent des domaines globulaires aux extrmits, encadrant une rgion riche en segments torsads (coiled-coil) qui serait implique dans la formation de dimres parallles. Lextrmit C-terminale de ces dimres serait ancre aux lments latraux, tandis que lextrmit N-terminale, oriente vers le centre de la structure, y raliserait la liaison entre les deux parties du complexe (i.e. le synapsis) via des interactions homotypiques ou en liant dautres protines de llment central, comme SYCE1 (Bolcun-Filas, Speed et al. 2009), SYCE2 (Bolcun-Filas, Costa et al. 2007) et TEX12 (Hamer, Wang et al. 2008) chez la souris. Les protines mises en jeu dans linitiation du synapsis sont assez bien identifies et caractrises chez S. cerevisiae. Selon les donnes les plus rcentes (Shinohara, Oh et al. 2008), Zip3 serait localise aux sites dinitiation du synapsis, y recruterait Zip1, puis les complexes Zip1-Zip3 intgreraient les protines Zip2, Zip3 et Spo16. Les protines Msh4 et Msh5 sont galement associes physiquement Zip3 (Agarwal et Roeder 2000). Lensemble de ces protines constitue 15

chez S. cerevisiae les complexes dinitiation du synapsis (CIS), partir desquels la protine Zip1 ralise progressivement le synapsis en formant les filaments transversaux. Les CIS dterminent la fois linitiation du synapsis et la diffrentiation des CO de type I. Ce couplage entre les deux processus nest apparemment pas retrouv chez tous les eucaryotes. Un homologue de Zip3 chez C. elegans (ZHP-3), est ncessaire la formation des CO mais nest pas impliqu dans le synapsis (Jantsch, Pasierbek et al. 2004). Chez A. thaliana, les CO de type I (interfrents, voir les sections 1.2.2.5.1. et 1.3.2.) sont abolis dans un mutant Atzip4, mais le synapsis est normal (Chelysheva, Gendrot et al. 2007).

1.2.1.1.3.5 Les entremlements de bivalents


Il arrive trs frquemment que des entremlements (interlocks) de bivalents se produisent au cours du zygotne quand un bivalent ou un axe est positionn entre deux axes aligns mais pas encore synapss. Ces structures peuvent tre rsolues de diverses manires (Holm et Rasmussen 1980), soit la cassure dun lment axial (c'est--dire de deux chromatides soeurs), soit le mouvement des chromosomes (Koszul et Kleckner 2009), soit encore par dsassemblage/rassemblage du CS (Kleckner et Weiner 1993). Ces mcanismes pourraient agir de faon concomitante (sujet revu par (Wang, Carlton et al. 2009)). On peut observer la rsolution dentremlements dans une grande varit despces comme le coprin (Holm, Rasmussen et al. 1981), le bombyx (Holm et Rasmussen 1980), lhomme (Rasmussen et Holm 1980), le seigle (Abirached-Darmency, Zickler et al. 1983), le mas (Wang, Carlton et al. 2009) et S. macrospora (Storlazzi, Gargano et al. 2010), entre la fin du zygotne et le dbut du pachytne. Il arrive cependant que les entremlements ne soient pas rsolus et persistent jusquen mtaphase (John 1976).

1.2.1.1.4 Pachytne
Au dbut du stade pachytne (du grec pakhys pais ), le synapsis est par dfinition achev. Cependant, au cours de la phase qui suit, des modifications de la morphologie des bivalents peuvent tre observes. En premier lieu, des variations de longueur du CS ont t notes au sein de diverses espces, sans quune rgle gnrale semble applicable tous les organismes. Ltendue de ces variations est parfois assez considrable, ainsi chez Schizophyllum les bivalents raccourcissent de 30% au cours du pachytne (Rasmussen et Holm 1980). Par ailleurs, les boucles formes au cours du zygotne par le synapsis de rgions inverses entre les homologues font lobjet dun processus d ajustement synaptique , qui permet de former un CS linaire au travers de ces rgions chez les mammifres (Moses, Poorman et al. 1982; Gabriel16

Figure 11. Nodules de recombinaison tardifs. Bivalent ZW de nandou au stade pachytne, color lacide phosphotungstique. Les nodules de recombinaison tardifs sont indiqus par des ttes de flche. La barre reprsente 1 m. (Pigozzi et Solari 1997).

DAPI

MLH1

superposition

Figure 12. Chiasma et foci MLH1. Sur ces clichs dun miocyte au stade diacinse chez A. thaliana, on observe les bivalents colors au DAPI gauche, les foyers MLH1 au milieu, et la superposition des images droite. La barre reprsente 10 m. (Chelysheva, Grandont et al. 2010).

Robez, Ratomponirina et al. 1986) mais apparemment pas chez les plantes (Loidl 1987; Anderson, Stack et al. 1988). Mais les modifications structurales les plus pertinentes pour ce qui concerne la recombinaison miotique ont lieu au niveau des nodules de recombinaison. Leur nombre dcrot beaucoup, leur forme, leur localisation et leur composition se modifient. Passe cette transition, les nodules sont qualifis de tardifs (Zickler et Kleckner 1999; Anderson et Stack 2005; Ashley 2008)). Les nodules tardifs ont une forme plus rgulire, souvent sphrique ou ovode, que les nodules prcoces ou de transition. Ils sont toujours associs la rgion centrale du SC (Schmekel et Daneholt 1998), contrairement certains nodules prcoces qui peuvent tre situs distance des axes non synapss au zygotne. Alors que les nodules prcoces sont prsents en trs grands nombre un peu partout le long des axes, les nodules tardifs sont relativement peu nombreux (Stack et Anderson 1986), et ont une distribution moins homogne. Les premires observations en microscopie lectronique de nodules tardifs ont rvl quils correspondaient trs bien en nombre comme en position aux chiasmas (Zickler et Sage 1981)(figure 11). La protine MLH1 a ensuite t identifie comme un marqueur des sites de crossing-over aux stades pachytne et diplotne, tout dabord chez lhumain et la souris (Barlow et Hulten 1998), puis dans de nombreuses espces eucaryotes comme le poulet (Pigozzi 2001), le poisson zbre (Moens 2006), A. thaliana (Franklin, Higgins et al. 2006) et la tomate (Lhuissier, Offenberg et al. 2007). La fonction miotique de MLH1 nest documente, ni chez C. elegans, ni chez D. melanogaster, bien que cette protine soit prsente dans les deux espces. Le marquage immunologique de MLH1 est aujourdhui communment utilis comme un outil de dtection prcoce des sites de crossing-over.

1.2.1.1.5 Diplotne et diacinse


Au cours du stade diplotne (grec diplo double) le complexe synaptonmal se dissocie peu peu et les homologues se sparent, sauf au niveau des chiasmas qui ne sont pas encore distinguables (voir par exemple chez lhomme (Holm et Rasmussen 1983), la souris (Lee, Iwai et al. 2003), A. thaliana (Ross, Fransz et al. 1996; Stronghill et Hasenkampf 2007)). Des clichs illustrant la disparition progressive des nodules tardifs et lapparition corrlative des chiasmas ont t obtenus chez le bombyx (Holm et Rasmussen 1980). Les chiasmas ne sont clairement visibles quune fois les chromosomes condenss, au cours de la diacinse, qui suit le diplotne. Par ailleurs chez A. thaliana, la protine MLH1 persiste au niveau de la plupart des chiasmas en diacinse (L. Chelysheva, sous presse, figure 12), ce qui constitue une preuve irrfutable de la correspondance entre NRT et chiasmas (de type I, voir plus bas).

17

Figure 13. Modle de la double jonction de Holliday, de Szostak et al. La rsection 5>3 dune cassure double-brin produit des extrmits 3 simple-brin. Lune dentre elles envahit un duplex homologue, ce qui produit une boucle de dplacement, laquelle la seconde extrmit de la cassure peut shybrider. La migration de branche de la demi jonction de Holliday ( gauche) permet, aprs synthse dADN et ligation, la formation dune double jonction de Holliday. En fonction de la position des sites de clivage, la structure est rsolue soit en deux crossovers, soit en un non-crossover. (Haber 2008).

1.2.1.2 Mtaphase, anaphase et tlophase


Le fuseau de division commence apparatre pendant la diacinse. Les fibres se connectent aux centromres des homologues au cours de la mtaphase. Puis le relchement de la cohsion entre chromatides surs au niveau des bras, mais pas au niveau des centromres permet la migration des homologues vers les ples opposs lanaphase (sujet revu par (Petronczki, Siomos et al. 2003)). La tlophase de premire division de miose est suivie dune interphase courte, pendant laquelle les chromosomes se dcondensent. La deuxime division miotique est assez rapide, moins de deux heures chez A. thaliana (Armstrong, Franklin et al. 2003), et aboutit la sgrgation des chromatides surs. Cet aspect de la miose ne sera pas prsent plus en dtails.

1.2.2 Description molculaire de la recombinaison miotique


Jusquaux annes 1980, les connaissances sur les mcanismes molculaires de la recombinaison miotique ont t dduites de donnes gntiques acquises chez des champignons tels que Neurospora crassa, Ascobolus immersus, S. pombe, et surtout S. cerevisiae. Puis les progrs de la biologie molculaire et de la biochimie ont contribu, essentiellement chez S. cerevisiae et S. pombe, lidentification des acteurs molculaires intervenant chacune des tapes. Trois modles de la recombinaison homologue ont marqu les progrs de la recherche dans ce domaine. Celui de Robin Holliday en 1964 (Holliday 1964), celui de Matthew Meselson et Charles Radding en 1975 (Meselson et Radding 1975), puis celui de Jack Szostack en 1983 (Szostak, Orr-Weaver et al. 1983). Ce dernier, baptis modle de rparation des cassures double-brin (RCDB), a fait office de paradigme jusquau dbut des annes 2000 (figure 13). Puis limplication en miose de voies alternatives de rparation par recombinaison homologue a t formellement dmontre (Allers et Lichten 2001; Allers et Lichten 2001; Hunter et Kleckner 2001). Cela a motiv une actualisation du modle de Szostak, qui demeure pour lessentiel vrai, dfaut dtre complet. Les donnes gntiques, et surtout molculaires, acquises chez les eucaryotes suprieurs (plantes et animaux) sont moins nombreuses. Cela est du plusieurs causes : il est actuellement toujours impossible de disposer dun nombre important de miocytes synchrones, ce qui complique lanalyse physique des intermdiaires de recombinaison et lanalyse biochimique des protines impliques. les cribles phnotypiques sont plus longs et la cration de souches mutantes est plus laborieuse et parfois infructueuse.

18

de nombreux gnes impliqus dans la recombinaison homologue miotique sont galement ncessaires la recombinaison somatique, et leur inactivation peut avoir des consquences pliotropes, ventuellement dltres, parfois ltales. Cependant des espces modles tels la souris, C. elegans et A. thaliana, ont dores et dj contribu trs significativement lavancement des connaissances sur les mcanismes de la recombinaison miotique, chacun en vertu de caractristiques propres permettant dtudier des aspects spcifiques (biochimiques, gntiques ou cytologiques) de la recombinaison miotique.

1.2.2.1 Formation des cassures double-brin


Ds 1909, Janssens postulait que la recombinaison miotique pouvait rsulter de la cassure suivie de la runion de segments de chromosomes homologues (Janssens 1909). La ncessit de cassures des chromosomes initiatrices du processus de recombinaison sest impose assez rapidement comme une vidence mais la preuve formelle de leur existence na t apporte qu'en 1989 (Game, Sitney et al. 1989; Sun, Treco et al. 1989). Limplication de la protine SPO11 dans la formation programme de cassures double-brin (CDB) de lADN, au cours de la prophase de miose, a t dmontre peu aprs (Cao, Alani et al. 1990; Bergerat, de Massy et al. 1997; Keeney, Giroux et al. 1997). Il sest avr par la suite que ce rle est conserv dans toutes les espces eucaryotes se reproduisant par voie sexue, notamment A. thaliana (Grelon, Vezon et al. 2001). SPO11 est effectivement responsable de la catalyse des CDB, mais nagit pas seule. Dautres protines sont ncessaires son activit, mais leur identit diffre selon les espces (Keeney 2001; Keeney 2008). Par ailleurs, la spcificit de coupure est contrle par des mcanismes encore assez mal connus. Ainsi chez S. cerevisiae, il est possible de cibler lactivit de Spo11 en la fusionnant, par exemple, au domaine de liaison lADN de GAL4 (Pecina, Smith et al. 2002; Robine, Uematsu et al. 2007). Rcemment, le rle de certaines modifications dhistones dans le contrle de la distribution des sites de CDB a t suggr chez S. cerevisiae (Borde, Robine et al. 2009), et surtout dmontr chez les mammifres (Baudat, Buard et al. 2010; Myers, Bowden et al. 2010; Parvanov, Petkov et al. 2010). Ce point sera trait plus en dtails dans la section 1.4.2.4.

1.2.2.2 Maturation des CDB


Une fois sa tche effectue, la protine SPO11 reste attache par une liaison phosphodiester (donc covalente) chacune des extrmits 5 de la CDB. Chez S. cerevisiae, elle est libre suite un clivage endonuclolytique de lADN, vraisemblablement mdi par Mre11 (dans le contexte du complexe MRX form avec les protines Rad50 et Xrs2) en collaboration avec Sae2 (Keeney 2008; Mimitou et Symington 2009),

19

Figure 14. De la formation de la cassure double-brin linvasion de brin. 1 Les extrmits 3 simple-brin sont formes grce aux protines MRE11 et EXO1, notamment. 2 Les extrmits sont recouvertes par RPA, ce qui limine les structures secondaires. 3 La protine RAD52 recrute RAD51 au niveau des extrmits recouvertes par RPA. 4 La formation du filament nucloprotique contenant RAD51 est mdie par RAD55 et RAD57. 5 Le filament nucloprotique trouve une squence dADN homologue. 6 RAD54 interagit avec RAD51 et permet lhybridation du brin invasif son complmentaire sur la chromatide homologue. (Krogh et Symington 2004).

de chaque ct de la CDB, ce qui gnre des extrmits simple brin 3 sortantes (figure 14). Puis, ces extrmits simple-brin sont rsectes davantage par un mcanisme encore incompltement lucid, faisant intervenir les protines Exo1, Sgs1 et Dna2 (Mimitou et Symington 2008; Manfrini, Guerini et al. 2010). Chez S. cerevisiae la longueur des tracts de rsection varie entre 1 et 2 kb (Borts et Haber 1989). Il semble que la cohsion des chromatides surs puisse la limiter (Klein, Mahr et al. 1999). Chez S. cerevisiae, les brins dADN gnrs par la rsection des extrmits de la cassure sont recouverts, mesure quils sont forms, par la protine RPA qui les empche de former des structures secondaires (Ehmsen et Heyer 2008). Puis des protines dites mdiatrices permettent le remplacement de la protine RPA par les recombinases Rad51 et Dmc1, qui sassemblent en hlice autour de chacune des extrmits simple-brin, formant des filaments nucloprotiques, ou nuclofilaments. Les protines mdiatrices sont bien connues chez S. cerevisiae. Pour Rad51, il sagit des protines Rad52 (Sung 1997), et des protines Rad55 et Rad57 (Sung 1997), qui sont des paralogues de Rad51 (figure 14). Les protines Mei5 et Sae3 sont des mdiateurs spcifiques de Dmc1 (Hayase, Takagi et al. 2004). Chez les animaux et les plantes, la situation est relativement moins claire. La protine RAD52 nest requise pour la recombinaison miotique ni chez la souris (Bannister et Schimenti 2004), ni chez A. thaliana (elle nexiste pas). Par contre la protine BRCA2 pourrait se substituer RAD52 dans ces deux organismes (Sharan, Pyle et al. 2004; Siaud, Dray et al. 2004; Dray, Siaud et al. 2006; Thorslund, Esashi et al. 2007). Par ailleurs il existe chez les mammifres cinq paralogues de RAD51, qui forment deux complexes distincts RAD51B/C/D-XRCC2 et RAD51C-XRCC3 (Liu, Tarsounas et al. 2007). Chez la souris, leur fonction miotique nest pas documente, les mutants prsentant une lthalit embryonnaire, bien que lactivit mdiatrice de RAD51B-C in vitro soit dmontre (Sigurdsson, Van Komen et al. 2001; Lio, Mazin et al. 2003). Chez A. thaliana, le complexe RAD51C-XRCC3 est ncessaire la recombinaison miotique (Bleuyard et White 2004; Bleuyard, Gallego et al. 2005).

1.2.2.3 Invasion de brin


Les nuclofilaments RAD51 et DMC1 possdent la proprit de pouvoir sassocier transitoirement un double brin dADN par des interactions dites paranmiques. Ces interactions leur permettent de dtecter , de rechercher par le jeu de collisions alatoires, une squence complmentaire avec laquelle ils peuvent former un intermdiaire (relativement) stable trois brins. Les liaisons paranmiques sont ensuite converties en liaisons plectonmiques 20

(i.e. entrelaces) entre le brin invasif et son complmentaire. Lchange de brins (sujet revu par (San Filippo, Sung et al. 2008)) homologues fonctionne progressivement, partir de lextrmit 3 du brin invasif. La structure molculaire rsultant dune invasion de brin extensive est appele boucle de dplacement (Hunter et Kleckner 2001). Chez S. cerevisiae, le processus dinvasion de brin est facilit par les protines Rad54 (figure 14), Rdh54/Tid1 (Shinohara, Shita-Yamaguchi et al. 1997; Schmuckli-Maurer et Heyer 2000; Shinohara, Gasior et al. 2000) et Mer3 (Mazina, Mazin et al. 2004). Un ou plusieurs homologues de Rad54 sont prsents chez les vertbrs (Bannister et Schimenti 2004; Wesoly, Agarwal et al. 2006), et les plantes (Osakabe, Abe et al. 2006; Klutstein, Shaked et al. 2008), mais bien que jouant un rle dans la recombinaison homologue somatique, ils nont pas de fonction miotique connue. Par contre, lhlicase MER3 est ncessaire, aussi bien chez A. thaliana (Mercier, Jolivet et al. 2005), le riz (Wang, Tang et al. 2009), que chez S. cerevisiae (Nakagawa et Kolodner 2002), la formation de la majorit des CO. De plus, dans toutes les espces o le gne dmc1 est prsent, le complexe des protines HOP2 et MND1 est galement impliqu dans la recombinaison homologue miotique. Chez S. cerevisiae et la souris, il possde in vitro une activit mdiatrice de linvasion de brin par les nuclofilaments Rad51 et/ou Dmc1 (Chen, Leng et al. 2004; Chi, San Filippo et al. 2007; Pezza, Voloshin et al. 2007). In vivo, en outre, le complexe HOP2-MND1 semble impliqu dans le contrle de la spcificit de linvasion de brin. En effet, linactivation des gnes Hop2 et/ou Mnd1 entrane la formation dinteractions non-homologues chez S. cerevisiae (Leu, Chua et al. 1998; Henry, Camahort et al. 2006), la souris (Petukhova, Pezza et al. 2005), et vraisemblablement A. thaliana (Vignard, Siwiec et al. 2007). Il est possible que MND1 et HOP2 collaborent plutt avec DMC1 que RAD51 chez S. cerevisiae et S. pombe (Saito, Tougan et al. 2004), mais les arguments en faveur de cette hypothse sont actuellement assez tnus. Alors que RAD54, RDH54, MER3, HOP2 et MND1 ont pour fonction premire de stabiliser le nuclofilament et/ou dtendre le duplex entre le brin invasif et son complmentaire, dautres protines seraient dotes dune fonction exactement inverse : les hlicases Mph1 (Prakash, Satory et al. 2009) et Srs2 (Dupaigne, Le Breton et al. 2008, 2008 #6690) de S. cerevisiae, BLM (Bachrati, Borts et al. 2006) et RTEL-1 (Barber, Youds et al. 2008 ) de mammifres, dsassemblent in vitro la boucle de dplacement. La fonction miotique de ces protines nest cependant pas bien lucide.

1.2.2.4 Le biais inter-homologues


Lors de la recombinaison homologue somatique, RAD51 est la seule recombinase constitutive des nuclofilaments. La protine DMC1, qui est code par un gne paralogue prsent dans la 21

majorit des espces ( lexception notamment de D. melanogaster et C. elegans), est implique spcifiquement pendant la miose. Sur un plan fonctionnel, ces deux recombinases ne sont pas redondantes et cela semble li une diffrence fondamentale entre cellules somatiques et miotiques. Dans le premier cas la rparation fait gnralement intervenir la chromatide sur, tandis que dans le second cas cest une chromatide homologue qui est le partenaire prfrentiel dans la plupart des espces (voir chez S. cerevisiae (Haber, Thorburn et al. 1984; Game, Sitney et al. 1989; Schwacha et Kleckner 1994)). Ce phnomne, dsign par lexpression biais interhomologue , est vraisemblablement de rgle chez les eucaryotes, lexception peut-tre de S. pombe. Dans cet organisme, la recombinaison miotique implique vraisemblablement deux chromatides surs plus frquemment que deux chromatides homologues (Cromie, Hyppa et al. 2006). Chez S. cerevisiae, Dmc1 et Rad51 sont toutes deux impliques dans la rparation par recombinaison homologue des CDB miotiques. Mais, tant donn que ces deux protines ne sont gure distinguables in vitro, aussi bien sur le plan de lactivit biochimique (invasion de brin) que sur celui de la structure des nuclofilaments (Sheridan, Yu et al. 2008), il semble raisonnable de supposer que ce sont leurs partenaires qui permettent de diffrentier leurs fonctions in vivo. De fait, il a t montr que les protine Mek1 et Hed1, inhibent la stimulation par Rad54 de lactivit dinvasion de brin de Rad51 (Busygina, Sehorn et al. 2008; Niu, Wan et al. 2009). En labsence de ces deux protines, Dmc1 nest plus ncessaire la rparation des CDB : Rad51 seule permet la recombinaison avec une chromatide homologue (Niu, Wan et al. 2009), bien quil soit possible que le biais homologue/sur soit abaiss. Cependant, la recombinase Dmc1 seule ne permet apparemment linvasion de brin que sur la chromatide soeur, pas sur une chromatide homologue, (Schwacha et Kleckner 1997). En outre, dans au moins une souche de S. cerevisiae (BR2495), labsence de Dmc1 ne provoque quune diminution et pas la quasi-disparition de la recombinaison impliquant une chromatide homologue (Rockmill, Sym et al. 1995). En somme, Rad51 est ncessaire la recombinaison avec une chromatide homologue, mais nest normalement pas suffisante, tandis que Dmc1 est suffisante la recombinaison avec la chromatide sur, mais nest pas ncessaire. Chez S. pombe, de mme que dans la souche BR2495 de S. cerevisiae, Dmc1 joue un rle relativement accessoire. En son absence, la recombinaison inter homologue est abaisse de 60 80%, mais toutes les CDB sont rpares et la viabilit des spores est quasi normale (Fukushima, Tanaka et al. 2000; Grishchuk et Kohli 2003; Grishchuk, Kraehenbuehl et al. 2004; Young, Hyppa et al. 2004). Dmc1 semble donc simplement faciliter la recombinaison entre chromatides homologues. Lhomologue de Rad51, Rhp51 joue par contre un rle crucial dans la rparation 22

des CDB miotiques quelle fasse intervenir une chromatide sur ou homologue (Grishchuk et Kohli 2003). Chez la souris, RAD51 est ncessaire la survie de lembryon (Lim et Hasty 1996), et des cellules qui en sont dpourvues prsentent une fragmentation chromosomique entranant lapoptose (Sonoda, Sasaki et al. 1998). Sa fonction miotique est inconnue mais elle est probablement essentielle la rparation des CDB programmes. Chez A. thaliana, labsence de RAD51 ninduit pas de phnotype vgtatif, mais un dfaut miotique trs marqu : les homologues ne synapsent pas et apparaissent fragments au stade diacinse (Li, Chen et al. 2004). Aussi bien chez la souris (Pittman, Cobb et al. 1998; Yoshida, Kondoh et al. 1998), que chez A. thaliana (Couteau, Belzile et al. 1999), la protine DMC1 semble ncessaire recombinaison entre chromatides homologues mais pas entre chromatides surs, dans la mesure o les individus mutants prsentent un dfaut de synapsis des homologues, mais pas de fragmentation chromosomique dtectable. Le mcanisme de diffrentiation des fonctions miotiques de RAD51 et DMC1 chez les animaux et les plantes est inconnu. Il se pourrait que la nature et lorganisation des recombinases contenues dans les nuclofilaments contribuent au mcanisme de mise en place du biais inter-homologue. En effet la co-localisation biochimique et cytologique de RAD51 et DMC1 (Tarsounas, Morita et al. 1999) chez la souris peut laisser penser que les deux protines forment des filaments nucloprotiques mixtes. Si cela est vrai, leurs proportions ou leur arrangement squentiel pourraient cependant diffrer entre les extrmits simple-brin. Lobservation par immunomarquage de foci non pas superposs mais adjacents chez S. cerevisiae (Shinohara, Gasior et al. 2000) suggre que les extrmits simple-brin pourraient lier exclusivement lune Rad51 et lautre Dmc1. Cette dernire hypothse semble dautant plus crdible que la maturation endo/exonuclolytique des extrmits (section 1.2.2.2.) ne procde pas symtriquement de part et dautres des CDB (Neale, Pan et al. 2005). Mme si les extrmits sont asymtriques, il semble que toutes deux puissent tre impliques concurremment dans un processus dinvasion de brin, bien quil soit possible que lun soit moins efficace que lautre et implique une chromatide sur plutt quune homologue (Oh, Lao et al. 2007).

1.2.2.5 Les voies de recombinaison homologue


Ltape dinvasion de brin constitue un carrefour du processus de recombinaison homologue. Les rgulateurs positifs ou ngatifs de lextension de la boucle de dplacement dterminent lvolution des intermdiaires de recombinaison. Dans la version classique du modle de RCDB, la boucle D est stabilise, et le brin dplac lie probablement RPA et Rad52, ce qui lui permet ensuite de shybrider la deuxime extrmit simple-brin (Nimonkar, Sica et al. 2009).

23

Figure 15. Voies de recombinaison homologue miotique. Les duplex dADN de deux chromatides homologues (i.e. non surs) sont reprsents en rouge et bleu. Les explications sont dans le texte. Daprs (Hollingsworth et Brill 2004).

Cette tape de capture de la deuxime extrmit oriente les intermdiaires vers la formation de CO. Toujours selon la version classique du modle de RCDB il existe ensuite une seule voie de formation de CO. Or, il est aujourdhui acquis quil en existe au moins deux, qui diffrent, aussi bien du point de vue de la nature des intermdiaires, que de celle des protines impliques.

1.2.2.5.1 Les crossovers de type I


Dans la premire voie, la continuit des brins est ensuite rtablie par synthse dADN et ligation. Ce processus engendre une structure stable possdant deux jonctions de Holliday (dJH). Daprs le modle de RCDB de Szostak, la rsolution des dJH peut donner naissance aussi bien des NCO qu des CO, selon la position des coupures endonuclolytiques au niveau des JH. En fait, les dJH produisent essentiellement des CO et pas ou trs peu de NCO (Allers et Lichten 2001). Il semble que cela pourrait nanmoins advenir par un processus de dissolution des dHJ faisant intervenir le complexe de protines Sgs1-Top3-Rmi1 chez S. cerevisiae (Ira, Malkova et al. 2003), ou son homologue chez les eucaryotes suprieurs : BLM-TOP3alpha chez les animaux (Plank, Wu et al. 2006; Raynard, Zhao et al. 2008), RECQ4A-TOP3alpha-BLAP75 chez les plantes (Hartung, Suer et al. 2007; Chelysheva, Vezon et al. 2008; Hartung, Suer et al. 2008). Chez S. cerevisiae, les gnes zip1, zip2, zip3, zip4, spo16, msh4, msh5 et mer3 (collectivement dnomms zmm) sont tous apparemment impliqus conjointement dans la diffrentiation des crossings-over de la premire voie et la formation du complexe synaptonmal (Lynn, Soucek et al. 2007). En effet, les phnotypes des mutants respectifs sont comparables, en premire approche (dfaut de formation du complexe synaptonmal, rduction du nombre de CO et noninterfrence des CO rsiduels). Nanmoins, des diffrences peuvent tre observes, qui laissent penser que les protines ZMM ne forment en fait pas un groupe homogne (Borner, Kleckner et al. 2004; Shinohara, Oh et al. 2008). Par ailleurs, le contrle coordonn de linitiation du synapsis et de la diffrentiation des CO de type I est observ spcifiquement chez S. cerevisiae. Chez A. thaliana par contre, ni ZIP4, ni MSH4, ni MSH5, ni MER3, ni SHOC1 (qui est un homologue prsomptif de ZIP2 (Macaisne, Novatchkova et al. 2008)) ne sont ncessaires au synapsis. Les protines MSH4 et MSH5 forment des htrodimres qui lient in vitro les structures dADN branches analogues aux intermdiaires dinvasion de brin et aux jonctions de Holliday (Snowden, Acharya et al. 2004; Snowden, Shim et al. 2008). In vivo, les htrodimres MSH4MSH5 sont prsums stabiliser les intermdiaires de recombinaison (les boucles de dplacement puis les dJH), en formant des sortes de pinces molculaires encerclant les chromatides homologues impliques dans la recombinaison (Hoffmann et Borts 2004; Ehmsen et Heyer 2008). Les htrocomplexes MSH4-MSH5 pourraient en outre restreindre laccs des complexes

24

de dissolution des dJH (Sgs1-Top3-Rmi1), et au contraire recruter les protines requises ultrieurement pour la maturation et/ou la rsolution des dJH (Hoffmann et Borts 2004). Les protines MLH1 et MLH3 ne font pas partie des ZMM, dans la mesure o, chez S. cerevisiae, elles ne sont pas ncessaires au synapsis. Nanmoins leur absence provoque galement une rduction du nombre de CO de type I, mais moindre que dans les mutants de gnes zmm, chez S. cerevisiae (Wang, Kleckner et al. 1999) et A. thaliana (Jackson, Sanchez-Moran et al. 2006). A cet gard, le mcanisme sous-jacent nest pas document, mais il est suppos que, de faon analogue MSH4-MH5, des htrodimres MLH1-MLH3 stabilisent localement les intermdiaires de recombinaison, une tape plus tardive de leur maturation (Argueso, Kijas et al. 2003; Hoffmann et Borts 2004). Il est assez vraisemblable quils puissent jouer dans ce cadre un rle dans le biais de rsolution des dJH en faveur des CO (Hoffmann et Borts 2004; Franklin, Higgins et al. 2006; Svetlanov, Baudat et al. 2008). Cependant chez S. cerevisiae, lobservation de profils de sgrgation post-miotique dans un mutant mlh1 (Wang, Kleckner et al. 1999; Argueso, Kijas et al. 2003) prouve que Mlh1 assure en outre une fonction de correction des msappariements prsents dans les htroduplex forms au cours du processus de recombinaison homologue. Cette fonction est assure en collaboration avec dautres facteurs, notamment Pms1, Mlh2, Mlh3, Msh2 et Msh6 (Wang, Kleckner et al. 1999; Stone et Petes 2006). Cet aspect de la recombinaison miotique est revu en dtails par (Surtees, Argueso et al. 2004). Il semble en fait que Mlh1 soit une sorte de pivot molculaire, permettant de recruter non seulement des protines impliques spcifiquement dans la correction des msappariements, mais aussi des protines telles que Exo1 et Sgs1 (Wang et Kung 2002; Dherin, Gueneau et al. 2009), peut tre dans le cadre dune coordination entre les activits de correction et de rsolution des intermdiaires de recombinaison.

1.2.2.5.2 Les crossovers de type II


La seconde voie fait intervenir les protines Mus81 et Mms4 (chez S. cerevisiae). MUS81 est une endonuclase de la famille XPF qui raliserait des coupures dcales des deux brins de la chromatide intacte. Chaque extrmit tant associe une extrmit diffrente de la chromatide invasive, la structure serait rsolue en CO par une synthse dADN suivie dune ligation (Gaskell, Osman et al. 2007). Selon un modle alternatif, le brin dplac par lextrmit invasive est cliv par MUS81, puis shybride au brin de lautre extrmit, ce qui aboutit, aprs synthse dADN et ligation la formation dune jonction de Holliday simple (Cromie, Hyppa et al. 2006)(figure 15). Les deux voies de formation de CO se distinguent par la nature des intermdiaires de recombinaison (dJH versus sJH), par lidentit des protines impliques, mais aussi et surtout par leur susceptibilit au phnomne dinterfrence. Les CO de la deuxime voie se forment

25

indpendamment les uns des autres, mais pas les CO de type I, qui ont tendance tre plus espacs quattendu du simple fait du hasard. Un ensemble de donnes biochimiques, cytologiques et gntiques, permet de conclure par des arguments directs (exprimentaux) ou indirects (fonds sur des analogies entre espces) que, dans la plupart des eucaryotes, la premire voie dtermine la majorit des CO : environ 90 % chez la souris (Guillon, Baudat et al. 2005), 85 % chez A. thaliana (Franklin, Higgins et al. 2006), 70 % chez la tomate (Lhuissier, Offenberg et al. 2007) et S. cerevisiae (Argueso, Wanat et al. 2004). La seconde voie produit par contre la totalit des CO chez S. pombe (Cromie, Hyppa et al. 2006) et peut-tre tous les CO indpendants de la premire voie dans les espces cites plus haut. Linterfrence des CO sera aborde dans la section 1.3.2.

1.2.2.5.3 Les non-crossovers


Les contributions relatives des protines facilitatrices (RAD54 et ses paralogues, MER3, HOP2MND1) ou antagonistes (BLM, SRS2, RTEL1) de la formation de la boucle de dplacement sont pour linstant mal connues, voire inconnues. Cependant, il est clair actuellement que la plupart des extrmits ne sont pas engages de faon stable dans un processus dinvasion de brin menant la formation de CO de type I ou de type II. La synthse dADN partir de lextrmit 3 du brin invasif permet, aprs dissolution de la boucle de dplacement naissante, la rhybridation au brin complmentaire au del de la cassure. La continuit de chaque brin, interrompue par les tracts de rsection, est rtablie par la synthse dADN. Ce mcanisme dhybridation de brin dpendante de la synthse dADN (HBDS) nengendre pas de CO, et ses produits sont donc gnralement baptiss non-crossovers (NCO) (Allers et Lichten 2001; Borner, Kleckner et al. 2004; Hunter 2007; McMahill, Sham et al. 2007)(figure 15).

1.2.2.5.4 Les intermdiaires de recombinaison aberrants


Il est thoriquement possible que les deux nuclofilaments puissent sengager dans un processus dinvasion de brin, pour shybrider chacun un brin complmentaire sur une chromatide sur ou homologue. Cela se produit effectivement, bien que rarement, chez S. cerevisiae. Les protines Sgs1 et Mus81 prviendraient la formation de tels intermdiaires impliquant plus de deux chromatides, qui sont susceptibles de produire des CO aberrants (Oh, Lao et al. 2007; Oh, Lao et al. 2008).

1.2.2.6 La rsolution des intermdiaires de recombinaison


Si le mcanisme de rsolution des sJH miotiques (intactes ou coupes) en CO de type II par le complexe Mus81-Mms4/Eme1 est actuellement bien dmontr (Gaskell, Osman et al. 2007), lidentit des protines responsables de la rsolution des dJH miotiques (en CO de type I) est

26

Figure 16. Le nombre de chiasmas par miose nest pas alatoire. Distributions observes (triangles) et attendues selon la distribution de Poisson (cercles) du nombre de chiasmas par miose dans un mutant Atmsh4 (haut) et un type sauvage (bas) chez A. thaliana. (Higgins, Armstrong et al. 2004).

par contre moins claire. Il est possible que leur identification formelle soit imminente. Des travaux rcents ont permis de rvler lexistence dune famille de protines conserve chez animaux et les champignons, dont larchtype est la protine Slx4 de S. cerevisiae, et qui stimule lactivit de nuclases structure-spcifiques, notamment MUS81. Lhomologue chez D. melanogaster (MUS312) fait partie dun complexe suspect de possder in vivo une activit de rsolution des dJH miotiques depuis 2002 (Yildiz, Majumder et al. 2002). Une tude rcente fournit des arguments en faveur dun rle similaire de lhomologue chez C. elegans (HIM-18). De plus, lactivit rsolvase de complexes comprenant MUS312 ou son homologue humain (BTBD12) a t dmontre in vitro (Andersen, Bergstralh et al. 2009; Munoz, Hain et al. 2009; Svendsen, Smogorzewska et al. 2009). Par ailleurs, la protine Yen1 de S. cerevisiae, et son homologue chez lhumain GEN1, sont trs clairement dotes dune activit de rsolution des jonctions de Holliday in vitro (Ip, Rass et al. 2008) Toutefois, Yen1 semble ne pas tre requise en miose pour rsoudre la plupart des dJH, mais agirait plutt en collaboration avec Mus81-Mms4 dans llimination dintermdiaires de recombinaison aberrants. La fonction miotique de GEN1 est pour linstant totalement inconnue.

1.3 Le nombre et la distribution des vnements de recombinaison miotique


1.3.1 Le nombre de crossings-over
1.3.1.1 La dtermination du nombre de crossings-over
Dans de nombreuses espces, le nombre de crossings-over forms le long dun bivalent nest pas une variable alatoire. Ainsi par exemple, chez la souris (Anderson, Reeves et al. 1999), A. thaliana (Higgins, Armstrong et al. 2004) ou la tomate (Sherman et Stack 1995; Lhuissier, Offenberg et al. 2007), on constate que la variance est beaucoup plus faible que celle dune distribution de Poisson de moyenne identique (figure 16). Ce phnomne est trs probablement universel, quelques exceptions prs, comme S. pombe ou A. nidulans, dans lesquelles le nombre de crossingsover est lev et varie sans doute alatoirement (Kohli et Bahler 1994). La caractristique la plus remarquable des distributions du nombre de crossings-over par bivalent est labsence de bivalents dpourvus de crossings-over. La quasi universalit de cette observation a permis de lriger en rgle : la loi du crossing-over obligatoire .

27

2 nombre de foci MLH1 1,5 R2= 0,9589 1

A
0,5 0 0 2 4 6 8 taille des bivalents (m) 10

2 nombre de foci MLH1 1,5 R2= 0,8187 1

B
0,5 0

10

12

14

16

taille des bivalents (m)


Figure 17. Relation entre taille physique des bivalents et nombre de crossings-overs. Ces graphes dcrivent la relation entre la taille physique moyenne absolue (en m) et le nombre moyen de crossings-overs par miose pour chacun des bivalents autosomaux chez la souris mle (A) et le chien mle (B). La droite de rgression (pointills) et le coefficient de dtermination R2 sont calculs pour les bivalents prsentant plus dun chiasma. Daprs (Froenicke, Anderson et al. 2002) et (Basheva, Bidau et al. 2008).

Le nombre de crossings-over est donc contrl, de sorte que chaque bivalent en possde au moins un. Quen est il au-del ? Dans des espces trs diverses on note une corrlation nette entre la taille des bivalents et le nombre de crossings-over, sans tenir compte du crossing-over obligatoire. Cette dernire prcision est dune importance cruciale : aussi courts que soient les bivalents, ils possdent toujours le crossing-over obligatoire, et inversement ils ne possdent de crossings-over supplmentaires quau-del dun seuil de taille donn. Les donnes acquises chez le chien (Basheva, Bidau et al. 2008) ou la souris (Lawrie, Tease et al. 1995) illustrent cela parfaitement (figure 17). Le nombre de crossings-over par unit de longueur des bivalents est une caractristique despce. Cette valeur est constante dun individu lautre ou dun sexe lautre, alors mme que la longueur absolue des bivalent peut varier, parfois de faon importante (leur longueur relative la longueur totale de CS est constante) (Kleckner, Storlazzi et al. 2003). Des donnes exprimentales sont depuis peu disponibles qui viennent conforter ces observations : chez C. elegans la mutation de gnes codant des sous-units de condensines entrane une augmentation corrle de la longueur des axes et du nombre de CO (Mets et Meyer 2009). La table 1 rcapitule les valeurs de densit de crossings-over par unit de longueur physique daxe chromosomique dans quelques espces. On constate une variation assez large entre le pigeon et le mas, par exemple. Il ne semble pas que la densit puisse tre relie ni au nombre de chromosomes, ni la taille du gnome. Par ailleurs, laugmentation de longueur des axes constate dans les mutants de sous-units de condensines chez C. elegans nest pas modifie quand les CDB miotiques sont limines (par mutation du gne spo11) (Mets et Meyer 2009). Ce dernier rsultat concorde avec ceux de travaux antrieurs mens chez S. cerevisiae, Coprinus cinereus et S. macrospora, qui montrent quun dfaut de formation des CDB miotiques ne parait pas affecter la formation des lments axiaux (Baudat, Manova et al. 2000; Celerin, Merino et al. 2000; Storlazzi, Tesse et al. 2003), et surtout (au moins dans les deux dernires espces) ne modifie pas non plus leur longueur. Ces donnes permettent donc a priori dcarter lhypothse dune dtermination de la longueur des axes par le nombre dvnements de recombinaison. On peut donc supposer linverse que cest la longueur des chromosomes en prophase de miose qui dtermine le nombre de crossings-over. Dans cette hypothse le contrle pourrait tre exerc indirectement, ds le stade leptotne, sur la formation des CDB partir desquelles se forment tous les crossings-over.

28

Espce 5 12 10 22 38 19 39 4,5 5,8 9,6 8,3-9,9 23,4 15,1 13,5

Longueur totale du CS (m)

Nombre de bivalents autosomaux

Longueur de CS (m) sources par crossing-over *

165

Arabidopsis thaliana Lycopersicon. esculentum Zea mays

230

(L. Giraut, comm. pers.; Lopez, Pradillo et al. 2008) (Sherman et Stack 1995) (Anderson, Doyle et al. 2003)

331

Homo sapiens

293-298

Canis lupus

246

(Sun, Oliver-Bonet et al. 2004; Codina-Pascual, Campillo et al. 2006) (Basheva, Bidau et al. 2008) (Froenicke, Anderson et al. 2002) (Pigozzi et Solari 1999)

Mus musculus

99

Columbia liva

248

* calcule en ne prenant en compte que les bivalents possdant plus d'un crossing-over

Table 1. Relation entre longueur du complexe synaptonmal et nombre de crossings-over.

1.3.1.2 La dtermination du nombre dintermdiaires de recombinaison


De mme que pour les crossings-over, le nombre dintermdiaires prcoces de recombinaison semble corrl la longueur des axes chromosomiques. Cela est observ par exemple pour les nodules prcoces chez le mas (Stack et Anderson 2002), ou pour les foyers RPA et MSH4 chez la souris (de Boer, Stam et al. 2006). De plus, chez C. elegans, la mutation de sous units de complexes condensine provoque corrlativement une augmentation de la longueur du CS et une augmentation du nombre de CDB (Mets et Meyer 2009). Le rapport entre les frquences de prcurseurs potentiels de CO (nodules prcoces ou foci RAD51) et de CO (sans tenir compte du CO obligatoire), par unit de longueur daxe varie entre 10 et 20 dans diverses espces de mammifres et de plantes (table 2). Le fait que ce rapport soit peu variable semble laisser penser quil pourrait exister un mcanisme conserv de slection des CO parmi les intermdiaires de recombinaison non diffrentis . A priori, il semble donc possible que les CDB interviennent alatoirement le long des chromosomes, que lune dentre elles volue en crossing-over obligatoire, et quune partie des autres engendre des crossings-over, par un mcanisme inconnu, ventuellement de slection alatoire. En fait ce scnario est faux, chez S. cerevisiae et vraisemblablement la plupart des eukaryotes. En 2006, une tude de mutants hypomorphes du gne spo11 chez S.cerevisiae montre que les CO sont soumis un phnomne dhomostasie. En effet, leur nombre est peu affect par la diminution du nombre de CDB (Martini, Diaz et al. 2006). En consquence, la seule variable explicative du nombre de crossings-over qui soit clairement identifie lheure actuelle est la longueur physique des axes chromosomiques. Dans cette perspective, rien ne laisse supposer que les crossings-over soient des vnements isols les uns des autres. De fait, les relations dinterdpendance entre crossings-over sont connues et tudies depuis les origines de la gntique moderne.

1.3.2 Linterfrence des crossings-over


1.3.2.1 Linterfrence gntique
En 1911 Morgan, convaincu que les gnes sont arrangs linairement le long des chromosomes, dcrit le crossing-over comme la cause de la sgrgation des facteurs gntiques. Il invoque les diffrences de distance physique entre les gnes pour expliquer les variations de la force de liaison entre les caractres (Morgan 1911). Par la suite, Sturtevant dresse la premire carte gntique dun segment de chromosome de drosophile (Sturtevant 1913), notant que les frquences de recombinaison pouvaient tre sous estimes du fait de lexistence de CO doubles. Cependant, il 29

Taille du gnome (Mb) 125 824 2365 2500 2390

Sources (Albini 1994; Lopez, Pradillo et al. 2008) (Anderson, Hooker et al. 2001) (Franklin, McElver et al. 1999; Anderson, Doyle et al. 2003) (Anderson, Reeves et al. 1999; Froenicke, Anderson et al. 2002) (Basheva, Bidau et al. 2008)

Longueur Nombre de Nombre de Ratio totale des NRP / crossings-over / crossing-overs Espce Nombre de NRP axes (m) longueur daxe longueur daxe * / NRP 0,08 18,8 Arabidopsis thaliana 234 156 1,5 0,066 14,5 Lycopersicon. esculentum 411 430,6 0,95 0,14 500 331 1,51 10,8 Zea mays 0,17 160 1,56 9,19 Mus musculus 250 0,105 Canis lupus 261 246 1,06 10,1 * calcule en ne prenant en compte que les bivalents possdant plus d'un crossing-over

Table 2. Relation entre longueur des axes et nombre de nodules de recombinaison prcoces.

observe galement que les CO doubles sont moins frquents quattendu si les CO taient indpendants les uns des autres. Quelque temps plus tard, ce phnomne dvitement mutuel est baptis interfrence, et mesur, pour deux intervalles gntiques donns, par le ratio entre le nombre de CO doubles attendus sans interfrence et le nombre effectivement obtenu , la valeur de cet index variant de 1 dans le cas de CO survenant indpendamment lun de lautre, + quand il ny a pas de CO doubles (Sturtevant 1915). Sturtevant note que la valeur du ratio diminue avec la distance sparant les intervalles, ce qui signifie que les CO deviennent indpendants quand ils sont suffisamment loigns. Mller introduit lanne suivant le terme de concidence, et dsigne aussi par ce terme le ratio inverse de lindex utilis jusque l par Sturtevant (Muller 1916). Le coefficient de concidence de Mller, C, est donc gal au rapport entre les frquences observe et attendue de CO doubles. Ce qui est attendu dans le cas o les CO surviennent indpendamment les uns des autres est en fait gal au produit des frquences de CO dans chacun des intervalles considrs. Linterfrence est alors dfinie comme I=1-C. Elle peut tre ngative, ce qui correspond des situations o la formation dun CO favorise la formation dun second CO. Elle est nulle pour deux CO indpendants lun de lautre. Linterfrence positive, variant entre 0 et 1, concerne tous les cas o les CO ont tendance sexclure lun lautre. Quand I=1, linterfrence est totale, et il nexiste pas de CO doubles. Le mode de calcul de C nest pas aussi trivial quil peut paratre au premier abord. En effet les individus prendre en compte dans chacune des catgories peuvent diffrer selon quils possdent ou non des CO ailleurs que dans les intervalles considrs pour noter les CO doubles (Weinstein 1918; Graubard 1932; Schweitzer 1934; Mather 1936; Stevens 1936). En 1938, Mather rcapitule plusieurs de ses contributions majeures ltude de linterfrence des CO (Mather 1938). En premier lieu, il introduit les notions distinctes dinterfrence chromatidienne et dinterfrence des crossings-over, qui se combinent pour aboutir linterfrence des CO observe dans les produits de miose. Linterfrence chromatidienne correspond la non indpendance entre les probabilits que les chromatides surs soient impliques dans des crossings-over sur un mme bivalent. De toute vidence les deux phnomnes sont susceptibles de diffrer quant leur base biologique. Il y a encore aujourdhui peu dvidences biologiques en faveur de lexistence de linterfrence chromatidienne, essentiellement parce que son tude nest possible que dans les organismes o les quatre produits dune miose peuvent tre isols (Lindegren et Lindegren 1942), si bien quelle est gnralement purement et

30

simplement ignore. Linterfrence des CO est donc suppose rsulter uniquement de linterfrence des crossings-over. Mather explicite le calcul des frquences de crossings-over par bivalent en fonction des frquences de CO par chromatide, sous lhypothse dune absence dinterfrence chromatidienne (voir aussi (Weinstein 1936)). Mather affirme quil existe une relation entre la taille cytologique des chromosomes et le nombre de chiasmas qui fait apparatre le chiasma obligatoire, et la proportionnalit dcrite dans la section 1.3.1.1. pour les chromosomes dont la taille excde ce quil nomme distance diffrentielle . Mather montre par ailleurs que la variance du nombre de crossings-over par bivalent nest pas compatible avec le processus de Poisson sous-jacent au modle de Haldane (Haldane 1919; Haldane 1931) et que cela implique un contrle de la distance entre deux crossings-over successifs le long des chromosomes. Il ne fournit pas de formulation mathmatique de cette distance dinterfrence mais propose quelle varie le long des chromosomes, ce qui selon lui explique labsence de corrlation subchromosomique entre distances gntiques et cytologiques observe chez la drosophile (section 1.3.3.1.) Enfin, Mather suggre que la formation des crossings-over procde squentiellement partir des centromres. Le concept de formation squentielle des crossings-over selon un processus de renouvellement se prte bien la formalisation mathmatique. Les processus de Poisson, qui postulent que la distribution des distances entre crossings-over suit une loi exponentielle, ne permettent pas de rendre compte de linterfrence (Haldane 1931). Dautres lois de distributions de distance ont donc t considres. Il est apparu que les lois gamma permettent de modliser simplement leffet de linterfrence sur la distance entre crossings-over successifs, dans le cadre dun processus de renouvellement stationnaire (Payne 1957; Cobbs 1978) (voir section 2.1). Il faut noter que ces modles supposent la stationnarit (i.e. que son niveau est constant une distance donne) de linterfrence dans une mtrique arbitraire et inconnue. Leur confrontation plusieurs jeux de donnes gntiques sest avre concluante (Foss, Lande et al. 1993; Foss et Stahl 1995; McPeek et Speed 1995; Zhao, Speed et al. 1995; Broman et Weber 2000; Lin, Cheng et al. 2001; Broman, Rowe et al. 2002), ce qui peut laisser supposer que linterfrence est stationnaire dans une mtrique gntique. Elle serait alors variable dans une mtrique physique, compte tenu de la non proportionnalit des distances gntiques et cytologiques une chelle sub-chromosomique (section 1.3.2.2.). Cependant, des analyses de concidence menes par exemple chez Podospora

31

anserina (Marcou, Masson et al. 1979) ou le bl (Saintenac, Falque et al. 2009) suggrent que linterfrence pourrait varier dans une mtrique gntique. Lexpression mathmatique de la distance sparant deux crossings-over fait intervenir une constante m appele paramtre de forme dont la valeur peut tre considre comme un index de la force de linterfrence. Certains des promoteurs de modles de type gamma ont mis en avant une justification biologique cette formalisation mathmatique en prtendant que m dsigne le nombre de NCO entre deux CO successifs le long des bivalents. Selon eux, il existerait un mcanisme de dcompte des cassures ou des intermdiaires de recombinaison qui permettrait de choisir ceux devant donner naissance des CO (Stahl, Foss et al. 2004). Il se trouve que la valeur de m calcule daprs les donnes de recombinaison dans divers organismes semble correspondre la balance observe entre le nombre dintermdiaires de recombinaison et le nombre de crossings-over. Ce modle de dcompte, combinant les avantages dune formalisation mathmatique relativement simple et dune bonne adquation aux donnes exprimentales, a donc sembl attractif, en dpit de labsence de preuves biologiques dun quelconque mcanisme sous-jacent. La dcouverte du phnomne dhomostasie des CO (voir la section 1.3.1.2.) a jet un certain discrdit sur ce modle.

1.3.2.2 Linterfrence cytologique


La facilit relative obtenir une abondance de donnes de recombinaison gntique dans une multitude despces a suscit de nombreuses analyses de linterfrence au travers de sa manifestation gntique, la concidence. Pourtant il ne parat a priori pas moins fond dtudier son expression cytologique, c'est--dire la distance physique entre crossings-over conscutifs. En effet, cela peut permettre dclairer la question de la mtrique pertinente, qui conditionne la fois le la lgitimit des analyses dadquation aux modles et la vraisemblance des hypothses biologiques concernant le mcanisme de linterfrence. Le progrs des techniques dobservation microscopique a permis, partir des annes 1960, de dcrire la distribution des chiasmas le long de chromosomes aux stades diplotne et diacinse. Ce qui tait simplement prsum jusque l est alors apparu clairement au travers de multiples tudes: la corrlation observe en miose entre les tailles gntique et physique des chromosomes nest pas valide une chelle subchromosomique (Henderson 1963; Southern 1967; Fox 1973; Hulten 1974; Shaw et Knowles 1976; Maudlin et Evans 1980; Laurie et Jones 1981; Hulten, Palmer et al. 1982; Laurie, Palmer et al. 1982; Saadallah et Hulten 1983; Laurie et Hulten 1985; Lawrie, Tease et al. 1995). Les tudes mentionnes ci-dessus montrent que les distances entre chiasmas conscutifs sont plus grandes quattendues sans interfrence. Il est galement remarqu que la distance entre

32

chiasmas semble corrle positivement la taille des bivalents, mais que par contre la distance minimale entre chiasmas ne varie pas selon les bivalents (Hulten 1974; Laurie et Hulten 1985; Lawrie, Tease et al. 1995). La dcouverte du complexe synaptonmal (Moses 1956; Moses 1968), puis la dmonstration de la correspondance entre nodules de recombinaison tardifs et chiasmas (section 1.2.1.1.4.) ont par la suite suscit des travaux visant tablir des cartes plus prcises de la distribution des crossingsovers le long des bivalents au stade pachytne (Sherman et Stack 1995; Pigozzi et Solari 1997; Pigozzi et Solari 1999; Anderson, Doyle et al. 2003). Cette prcision accrue des mesures atteste chez la tomate (Sherman et Stack 1995) que la distance moyenne absolue entre crossings-over est corrle la taille (euchromatinienne) du SC. Cest lidentification de la protine MLH1 comme un marqueur des sites de crossing-over qui a permis dobtenir des donnes la fois suffisamment nombreuses et prcises pour envisager des analyse prcises de linterfrence cytologique . Ces approches sont plus appropries que toutes celles qui ont pu les prcder dans la mesure o la protine MLH1 est cense ne marquer que les futurs crossings-over de type I, soumis linterfrence, alors que les nodules tardifs ou les chiasmas identifis en diacinse constituent une population mixte de type I et de type II, insensibles linterfrence (section 1.2.2.5.2.). La premire tude de ce genre, mene chez la souris (Anderson, Reeves et al. 1999), semble tablir que seule la distance relative entre foci MLH1 (en % de la longueur euchromatinienne de chaque SC) est invariante dun bivalent lautre. Nanmoins, les donnes dun article publi trois annes plus tard rvlent en fait, toujours chez la souris (Froenicke, Anderson et al. 2002), que la distance absolue entre foci MLH1 est effectivement corrle positivement la taille du SC, mais que la distance relative est par contre corrle ngativement. Cela est confirm par une tude ralise ultrieurement sur lhumain (Codina-Pascual, Campillo et al. 2006). Malgr tout, des analyses dinterfrence cytologique ont par la suite t ralises chez la souris (de Boer, Stam et al. 2006; de Boer, Dietrich et al. 2007), lhumain(Oliver-Bonet, Campillo et al. 2007 ; Lian, Yin et al. 2008 ; Ferguson, Leung et al. 2009), ou la tomate (Lhuissier, Offenberg et al. 2007), qui font usage de la distribution gamma pour estimer lintensit de linterfrence partir de distances entre foci MLH1 exprimes en pourcentage de la longueur des bivalents. Ladquation des donnes au modle de distribution gamma est juge satisfaisante dans tous les cas.

33

1.3.2.3 Les variations de lintensit de linterfrence


Sans prjuger du bon espace descriptif de linterfrence, il existe aujourdhui un consensus (relatif) pour considrer que le paramtre de forme des distributions gamma, qui permet de rcapituler soit des donnes de distance physique ou gntique, entre crossings-over ou CO respectivement, soit des donnes de frquence de CO doubles (via un traitement mathmatique appropri), peut tre utilis comme un index de la force , de lintensit de linterfrence. Tirant parti de cet a priori, une srie de travaux dcrivant des variations prsumes de lintensit de linterfrence, intra- ou inter-chromosomique, ou entre les sexes, ou encore entre individus, ont t publis depuis une dizaine dannes (Broman et Weber 2000; Lin, Cheng et al. 2001; Broman, Rowe et al. 2002; Lam, Horn et al. 2005; Solignac, Mougel et al. 2007; Lian, Yin et al. 2008). Les conclusions de ces tudes sont gnralement assez peu solides. Compte tenu de la difficult de ces analyses, lie notamment la taille des chantillons, aux ventuels biais, au choix des mthodes de traitement mathmatique et statistique, cela ne parat a priori pas trs tonnant. De plus, des variations inter-chromosomiques de la proportion de crossings-over de type II, insensibles linterfrence, peuvent a priori effectivement entraner des fluctuations de linterfrence apparente mesure partir de donnes gntiques (Lam, Horn et al. 2005). Dans les espces o les crossings-over de type II sont trs frquents, comme la tomate (30 %), linterfrence entre foci MLH1 (correspondant aux crossings-over de type I) est plus forte quentre nodules de recombinaison tardifs (correspondant tous les crossings-over) (Lhuissier, Offenberg et al. 2007). Il existe malgr tout au moins une tude montrant une variation inter-chromosomique de lintensit de linterfrence entre foci MLH1, c'est--dire entre crossings-over interfrents (Lian, Yin et al. 2008).

1.3.2.3.1 Y a-t-il une mtrique descriptive de linterfrence ?


Alors que les partisans du modle de comptage affirment que seule la mtrique gntique est adquate la modlisation de linterfrence (Housworth et Stahl 2009), les cytologistes utilisent les distributions gamma comme un simple outil, sans toujours considrer la pertinence de leurs analyses. Pour ajouter encore un peu la confusion, plusieurs travaux danalyse fine de linterfrence ont t raliss en considrant une mtrique physique de longueur dADN en pb (Chen, Tsubouchi et al. 2008; Mancera, Bourgon et al. 2008). La question du bon espace descriptif de linterfrence est donc toujours ouverte.

34

100 %

=> 2 chiasmas

100 %

=> 1 chiasma

50 %

25 %

=> 1,5 chiasmas

25 %
Figure 18. Fusion chromosomiques et interfrence des crossings-over chez C. elegans. A Deux paire dhomologues (lune rouge, lautre bleue), forment chacune un crossing-over par miose dans un type sauvage. B Dans un homozygote pour la fusion des chromosomes, le bivalent ne forme quun seul crossing-over. C Dans un htrozygote, le bivalent forme en moyenne 1,5 crossing-over. Adapt de (Hillers et Villeneuve 2003).

Cependant, plusieurs tudes rcentes dmontrent que les tentatives de modlisation, quand elles sont confrontes des donnes exprimentales de bonne qualit, peuvent savrer fructueuses. Ainsi, des modles tenant compte du crossing-over obligatoire ou des crossings-over de type II ont t labors, qui permettent de dcrire plus fidlement la ralit biologique, et dmettre des prdictions exprimentalement testables (Falque, Mercier et al. 2007; Falque, Anderson et al. 2009). Notamment, l dernire tude cite questionne la pertinence de lusage des distributions gamma en comparant leur adquation un jeu de donnes exprimentales celle dune simulation base sur un modle de tension mcanique (voir section 1.3.2.3.2.).

1.3.2.3.2 Les modles biologiques de linterfrence


On considre volontiers linterfrence des crossings-over comme un signal manant dintermdiaires de recombinaison dsigns (comme de futurs crossings-over), et se propageant le long des bivalents pour rprimer la dsignation des intermdiaires voisins. Effectivement, une tude publie en 2003 suggre cela prcisment: chez C. elegans, des lignes htrozygotes pour la fusion des chromosomes IV et X forment en moyenne 1,5 crossing-over sur lensemble de ces chromosomes, alors que les lignes homozygotes pour les chromosomes fusionns et non fusionns forment respectivement 1 et 2 crossings-over (Hillers et Villeneuve 2003)(figure 18). En 2004, un modle biologique de linterfrence des crossings-over reposant sur la transmission dune contrainte physique le long des paires de chromosomes homologues a t propos (Kleckner, Zickler et al. 2004). Il peut tre rsum ainsi : An intrinsic effect of stress-promoted buckling will be local relief of axial compression stress at the affected site. That local stress relief (relaxation) will then spread outwards in both directions, thereby disfavoring subsequent buckling, et thus CR designation events, at adjacent bridges. This effect would explain interference (Borner, Kleckner et al. 2004). Ce modle nest pas le premier du genre. Ds 1916, Mller imaginait une explication analogue: ..breakage might be thought of as resulting from the tightness of the twisting, for then a breakage of the threads at one point would relieve the tension of the filaments for some distance along the line et so tend to prevent another breakage from occurring near by.. (Muller 1916). Mather, reprenant les observations de Darlington sur lenroulement relationnel (Darlington 1935) propose en 1938 un mcanisme similaire : Coiling stresses are relieved in this way by crossing-over. The chromosomes show longitudinal cohesion et in consequence reduction of the stress occurs for some distance on either side of the chiasma. This naturally reduces the chance of crossing-over in these regions et results in interference .

35

Figure 19. Foci MLH1 au stade pachytne. Dans ce spermatocyte humain au stade pachytne, la cohsine SYN1 des axes est marque en rouge, les centromres en bleu, et les foci MLH1 en jaune. (Sun, Oliver-Bonet et al. 2004).

Dautres types de signal sont envisageables. Par exemple une raction de polymrisation se propageant le long des bivalents. Il a t propos initialement que ce mcanisme concide avec la mise en place du complexe synaptonmal (Egel 1978). Les NRP (prsents au zygotne) et NRT (prsents en pachytne) ont t interprts dans ce cadre respectivement comme les substrats et les produits de linterfrence (Rasmussen et Holm 1978). Quelques annes plus tard, une simulation informatique base sur un modle de polymrisation a t labore pour rendre compte des donnes exprimentales (King et Mortimer 1990). Le rle du CS dans linterfrence a rcemment t mis en doute quand il sest avr que les CIS chez S. cerevisiae manifestaient de linterfrence en labsence de la protine Zip1, donc sans synapsis (Fung, Rockmill et al. 2004). Le CS tant cependant ncessaire la formation des CO de type I, il permet limplmentation de linterfrence sans toutefois en en constituer le vecteur. Les modles biologiques de linterfrence, quils soient bass sur la transmission dune contrainte mcanique, ou sur la propagation dune raction de polymrisation, sont certes attractifs sur un plan conceptuel, mais ils ne sont pas tays par des arguments exprimentaux trs solides. A lheure actuelle, aucune activit molculaire contrlant spcifiquement linterfrence, i.e. qui ne soit galement ncessaire la formation des crossings-over de type I, na pu tre identifie.

1.3.3 La distribution des crossings-over


1.3.3.1 Le dterminisme chromosomique de la localisation des crossings-over
Ds les annes 1930, la comparaison entre les cartes gntiques et cytogntiques (de chromosomes polytnes) de drosophile (Painter 1934; Bridges 1935; Mather 1938) suggrait que les crossings-over ne sont pas distribus uniformment le long des chromosomes. Les premires observations cytologiques de localisation des chiasmas sont anciennes, mais imprcises. Il a fallu attendre les annes 1960 pour que soient publies des tudes de localisation fine des chiasmas le long de chromosomes aux stades diplotne et diacinse (section 1.3.2.2.). La prcision des cartes de distribution de chiasmas a t par la suite assez largement dpasse par les tudes de localisation de NRT le long du complexe synaptonmal. Depuis un peu plus de dix ans, cest limmunolocalisation de la protine MLH1 qui est communment utilise pour dterminer la position des crossing-overs le long des chromosomes au stade pachytne (figure 19). Des cartes de distribution de foci MLH1 ont t dresses dans des espces aussi diverses que la souris (Anderson, Reeves et al. 1999), lhumain mle (Sun, Oliver-Bonet et al. 2004; Codina-Pascual, Campillo et al. 2006; Sun, Oliver-Bonet et al. 2006; Oliver-Bonet, Campillo et al. 2007; Ferguson, Leung et al. 2009) ou femelle (Tease, Hartshorne et al. 2002), la tomate (Lhuissier, Offenberg et al. 36

Figure 20. Relation entre les distributions de nombre et de position des crossingsover sur les bivalents. a) Distribution globale des foci MLH1 chez deux humains mles (C6 et C7) le long du bivalent n8 au stade pachytne.. b) Distributions des foci MLH1 chez C7 et C8 le long du bivalent n8, selon quil possde un deux ou trois foci. (Codina-Pascual, Campillo et al. 2006).

2007), la musaraigne (Borodin, Karamysheva et al. 2008), le vison (Borodin, Basheva et al. 2009), le chien (Basheva, Bidau et al. 2008) le diamant mandarin (Pigozzi 2008) le poisson zbre (Kochakpour et Moens 2008), etc. Il ressort en premier lieu de ces travaux que la localisation des crossings-over le long des bivalents nest pas homogne. Au contraire, ils rvlent clairement une forte variation de la frquence de recombinaison par unit de longueur physique. Un examen systmatique des profils de distribution fait apparatre, de faon constante, que les tlomres, les rgions pri-centromriques et les organisateurs nuclolaires (constitus de rptitions dADN ribosomique) sont dpourvus de crossings-over. Ensuite, dans de nombreuses espces, la longueur du complexe synaptonmal varie entre les sexes, parfois de manire trs importante (Lynn, Koehler et al. 2002; Wallace et Wallace 2003; Tease et Hulten 2004; Kochakpour et Moens 2008), et le nombre de crossings-over varie corrlativement. Par ailleurs, pour un bivalent particulier, contenant un nombre donn de crossings-over suprieur un, les cartes de distribution tablies chez lhumain femelle (Tease, Hartshorne et al. 2002) ou mle (Sun, Oliver-Bonet et al. 2004; Codina-Pascual, Campillo et al. 2006), ou chez la souris (Lawrie, Tease et al. 1995; Anderson, Reeves et al. 1999) rvlent une alternance de creux et de pics de grande amplitude, qui sexplique manifestement par le contrle quexercent conjointement linterfrence et la taille des bivalents sur la position relative des crossings-over multiples. Par contre, si tous les crossings-over sont considrs, quel que soit leur nombre, les profils se superposent, de sorte que la variation de frquence le long du bivalent a tendance paratre plus stochastique. Cet effet est dautant plus marqu que la variance du nombre de crossings-over est importante, donc pour les bivalents les plus longs. De plus, il est tabli que la taille des bivalents peut varier beaucoup, la fois entre les cellules dun mme individu (Lynn, Koehler et al. 2002 ; Codina-Pascual, Campillo et al. 2006 ; Kochakpour et Moens 2008) et entre individus dune mme espce (Sun, Trpkov et al. 2005). Cette variation de taille augmente la variance du nombre de crossings-over et/ou modifie leur localisation, si bien que leffet de superposition de profils htrognes est encore accru. Ceci est bien illustr par lexemple de la figure 20. En bref, les cartes de distribution font apparatre, quand il existe une proportion notable de bivalents possdant plus dun crossing-over, une lvation de frquence prs des extrmits que lon peut imputer (au moins en partie) laction de linterfrence, et des fluctuations irrgulires de la frquence des crossings-over interstitiels, qui nont pas de localisation moyenne prfrentielle. Les crossings-over uniques, prsents essentiellement sur les

37

chromosomes courts, semblent plutt situs au milieu des bras (chez la souris (Lawrie, Tease et al. 1995) et lhumain femelle (Tease, Hartshorne et al. 2002)). Il convient de noter que dans les travaux cits ci-dessus, la position des crossings-over, des NRT ou des foci MLH1 est invariablement exprime relativement la taille des bivalent sur lesquels ils ont t observs. On peut imaginer que cela introduit un biais dinterprtation des rsultats, compte tenu de la variation de taille des bivalents voque plus haut (section 1.3.1.1.), mais il ne semble pas possible de procder autrement. Il est difficile de dgager une quelconque rgle, ou mme une simple tendance, sappliquant la distribution des crossings-over dune manire universelle, c'est--dire dans les deux sexes, et dans toutes les espces. Nanmoins, certaines particularits des profils de distribution de crossingsover ne semblent pas explicables par la seule action de linterfrence. Par exemple chez lhumain mle, la frquence des crossings-over atteint un pic trs net prs de lextrmit de certains chromosomes, puis chute brusquement dans la rgion la plus distale (Codina-Pascual, Campillo et al. 2006). Il se pourrait donc que la localisation des crossings-over manifeste une prfrence intrinsque pour certaines portions des chromosomes, qui pourrait, par exemple, tre dtermine par le profil dinitiation du synapsis, comme cela a parfois t propos (Jones 1984; Stack et Anderson 2002; Lynn, Ashley et al. 2004; Anderson et Stack 2005; Gaut, Wright et al. 2007).

1.3.3.2 Le dterminisme gnomique de la distribution des crossingsover


Sil y peu dvidences dune dtermination chromosomique per se de la distribution des crossingsover, il parat lgitime de sinterroger sur lexistence dun contrle sub-chromosomique rgional li la composition du gnome. Pour au moins deux raisons distinctes. La premire est dordre volutif : les CO gnrent de nouvelles combinaisons gntiques ; en consquence leur position vis--vis des gnes, et plus gnralement des lments gntiques considrs comme actifs , peut ne pas tre slectivement neutre. En second lieu, dans un rfrentiel physique la frquence rgionale de CO nest pas homogne, quelle que soit lchelle considre, dans la plupart des espces. La croissance exponentielle de leffort de squenage systmatique des gnomes a permis depuis environ dix ans dentreprendre ltude, une chelle sub-chromosomique, des corrlations entre le taux de recombinaison et divers paramtres gnomiques, tels que la densit en gnes, en pseudo-gnes, en lments transposables, en rptitions de divers types, le contenu en G et C (GC%), la fraction CpG, etc.

38

Lune des premires tudes de ce type, ralise en 2002 sur le gnome humain (Kong, Gudbjartsson et al. 2002), rvle que la frquence de recombinaison est corrle essentiellement avec le GC% et la fraction CpG. Curieusement, la corrlation positive avec le GC% mise en vidence par rgression simple est inverse quand elle calcule en rgression multiple, c'est--dire quand la fraction CpG est galement prise en compte. Autrement dit, il semble que la recombinaison est accrue dans les rgions riche en G et C, mais quelle lest encore plus dans les rgions pauvres en G et C dont la fraction CpG est leve. Les autres facteurs analyss, notamment la densit en gnes, sont trs faiblement corrls avec le taux de recombinaison. Par la suite, des tudes similaires menes sur la souris et le rat (Jensen-Seaman, Furey et al. 2004; Shifman, Bell et al. 2006), labeille (Beye, Gattermeier et al. 2006) et A. thaliana (Drouaud, Camilleri et al. 2006) ont galement montr que, parmi des corrlations toutes assez faibles, le GC% et/ou la fraction CpG sont les meilleures variables explicatives du taux de recombinaison. Chez le rat, la souris et A. thaliana, conformment ce qui a t observ chez lhumain, les rgions pauvres en G et C dont la fraction CpG est leve prsentent une augmentation du taux de CO. Chez le poulet, le taux de CO est trs fortement corrl au GC% (Duret et Galtier 2009). Ces rsultats peuvent paratre relativement peu persuasifs, bien quil soient concordants. Cependant, ils ont t obtenus par lanalyse dintervalles gnomiques de grande taille (125 kb chez labeille, 200 kb chez A. thaliana, 680 kb chez le poulet, 3 Mb chez lhumain). Or, des tudes ralises ultrieurement sur le dsquilibre de liaison dans des populations humaines, trs petite chelle (1 kb), laide doutils mathmatiques labors, ont rvl quil existe en fait des corrlations nettes, mais dont la force est trs dpendante de lchelle danalyse (Myers, Spencer et al. 2006; Spencer, Deloukas et al. 2006). Ainsi le GC% est assez nettement corrl positivement avec le taux de recombinaison, mais sur des intervalles de 8 512 kb, peut tre mme sur des intervalles de taille distincte dans cette gamme. Le contenu en exons est par contre corrl ngativement sur des chelles de 100 500 kb. La signification de la corrlation apparemment universelle entre le taux de recombinaison et le contenu en G et C a et fait encore lobjet dun dbat. Pour certains auteurs, la recombinaison entranerait une augmentation du %GC via un phnomne de conversion gnique favorisant les G et C (CGFGC) (Meunier et Duret 2004; Dreszer, Wall et al. 2007; Duret et Arndt 2008; Duret et Galtier 2009). Le CGFGC dsigne un biais de composition introduit lors de la conversion gnique (mitotique ou miotique) par la rparation prfrentielle des msappariements A:G, T:C, A:C, ou T:G en appariements G:C plutt que A:T (Birdsell 2002). Selon lopinion inverse, le CGFGC ne semble pas susceptible de produire un accroissement notable du %GC, compte tenu 39

Figure 21. Le biais de transmission alllique du la conversion gnique. A Reprsentation schmatique des chromatides impliques dans un vnement de CO un point chaud. Les CDB ont lieu seulement sur une chromatide du parent 1 (noire). CO1 et CO2 reprsentent les crossovers rciproques issus dune CDB. B Distribution des points dchange des CO1 (en rouge), des CO2 (en bleu) et du cumul (en noir). Les deux courbes se chevauchent en raison de la dispersion des vnements de CDB. La flche double reprsente la longueur moyenne des tracts de conversion gnique associs aux CO C Taux de transmission des allles du parent 2 parmi les CO. D Densit de conversion gnique rsultant des vnements de NCO. (Baudat et de Massy 2007a).

de la localisation trs troite et de la volatilit volutive de la frquence de recombinaison une chelle locale (celle du point chaud, voir section 1.4.). Le %GC serait dans ce cas plutt un facteur explicatif , mais il reste alors expliquer comment. La CGFGC est parfois (malencontreusement) confondue (Coop et Myers 2007; Mancera, Bourgon et al. 2008) avec le biais de transmission alllique du la conversion gnique (BCG) qui dsigne simplement un sens de conversion gnique (i.e. dun allle vers un autre) prfrentiel, rsultant des variations interallliques de la frquence dinitiation de la recombinaison (section 1.4.2.2.)(figure 21). Le BCG est suppos tre lorigine dun phnomne distinct, le paradoxe du point chaud , dont il sera question dans la section 1.4.2.4. Il ny a pas dvidence que le BGC soit reli, ni comme cause, ni comme consquence, aux variations du %GC.

1.3.4 La distribution des cassures double-brin


Ni les CDB chez S. cerevisiae (Baudat et Nicolas 1997; Gerton, DeRisi et al. 2000; Blitzblau, Bell et al. 2007; Buhler, Borde et al. 2007), ni les foci gamma H2AX (qui marquent les lieux de rparation des CDB miotiques) chez la souris (Grey, Baudat et al. 2009), ni les NRP chez le mas (Stack et Anderson 2002), ne sont distribus de faon homogne le long des chromosomes. Existe-t-il alors une relation entre les distributions dvnements de recombinaison prcoces et les distributions de crossings-over ? Effectivement, les variations de frquence des NRP et des NRT le long du CS sont peu prs parallles chez le mas (Stack et Anderson 2002). Il en va de mme des distributions de foci RPA1 et MLH1 chez lhumain mle (Oliver-Bonet, Campillo et al. 2007). Mais chez la souris par contre, on nobserve aucune corrlation entre la distribution des foci gammaH2AX, et celle des foci MLH1 le long du chromosome 17 (Grey, Baudat et al. 2009). Chez S. cerevisiae, les distributions des CO et des NCO le long du chromosome 1 sont galement trs discordantes (Mancera, Bourgon et al. 2008). Il semble donc assez douteux que les distributions inhomognes de crossings-over lchelle des chromosomes rsultent principalement du dterminisme de la localisation des CDB. Cependant, plusieurs tudes dmontrent par ailleurs que les vnements de recombinaison prcoces ne sont pas indpendants les uns des autres. Plus prcisment, ltude de la distance entre vnements successifs rvle quils sont sujets une interfrence significative. Cela est vrai des NRP dans diverses espces de plantes (Anderson, Hooker et al. 2001), des foci MSH4 chez la souris (de Boer, Stam et al. 2006) et RPA chez lhumain mle (Oliver-Bonet, Campillo et al. 2007), des foci Mer3 et Msh4 chez S. macrospora (Storlazzi, Gargano et al. 2010). Chez S. cerevisiae, une telle analyse de linterfrence cytologique nest gure faisable, mais lexamen des distances physiques (en pb) sparant les produits de recombinaison des diffrentes classes, mesures au

40

sein de ttrades, montre que les NCO interfrent avec les CO, bien quelles ninterfrent pas entre elles (Mancera, Bourgon et al. 2008). Ces donnes rcentes dmontrant lexistence dun niveau dinterfrence prcoce entre intermdiaires de recombinaison peuvent tre rapproches de rsultats plus anciens, qui dcrivent un effet de comptition entre sites de CDB chez S. cerevisiae, agissant sur des distances courtes et moyennes (jusqu 60 kb) (Wu et Lichten 1995; Xu et Kleckner 1995; Fan, Xu et al. 1997; Ohta, Wu et al. 1999; Robine, Uematsu et al. 2007). Ce phnomne dhomostasie rgionale des CDB pourrait sexpliquer par la liaison comptitive dun facteur diffusible prsent en quantit limitante le long des axes.

1.4 Les points chauds de recombinaison miotique


Les travaux de cytologie raliss chez les eucaryotes suprieurs ont permis dtablir formellement que les crossings-over ne sont pas localiss de faon homogne le long des chromosomes. Mais le niveau de rsolution de ces tudes est limit, et ne permet pas de dcrire la distribution des vnements de recombinaison trs petite chelle. Ce sont les tudes gntiques menes chez les champignons qui ont rvl, partir des annes 1960 et 1970, que la recombinaison miotique est en fait localise dans des zones trs discrtes des gnomes. Cette observation a par la suite t tendue aux eucaryotes dotes de gnomes volumineux, de sorte quun modle unitaire de distribution des vnements de recombinaison miotique est apparu au dbut des annes 1990 (Lichten et Goldman 1995). Ce modle de points chauds de recombinaison sapplique diverses espces de champignons, danimaux et de plantes, mais il nexiste pour linstant pas de preuve quil soit universellement valide : aucun point chaud de recombinaison miotique na t dcrit ni chez C. elegans, ni chez D. melanogaster, en dpit de labondance de travaux raliss sur ces deux organismes.

1.4.1 Les points chauds de recombinaison miotique chez les champignons


De nombreuses tudes consacres la recombinaison miotique dans diverses espces de champignons ont rvl que la frquence de recombinaison gnique est localement dpendante de la position. On observe trs gnralement une variation polarise bidirectionnelle partir dun site localis, selon les premiers travaux publis, prs de lextrmit 5 des gnes. Cest le cas notamment pour les locus me-2 chez N. crassa (Murray 1963), b2 chez A. immersus (Leblon 1972), M26 chez S. pombe (Goldman et Smallets 1979), ARG4 (Nicolas, Treco et al. 1989; Sun, Treco et al. 1989; Schultes et Szostak 1990) et HIS4 (Detloff, White et al. 1992) chez S. cerevisiae. Dans le 41

cadre des premiers modles de la recombinaison homologue, notamment celui de R. Holliday (Holliday 1964), cela suggrait lexistence de sites dinitiation -par des cassures de lADNdiscrets, c'est--dire prsents dans des zones troites et spcifiques des gnomes (Haber 2007). Cela paraissait tre en outre une caractristique assez gnrale de la recombinaison miotique chez les champignons. Cette hypothse a t valide ultrieurement, tout dabord par la dmonstration que des CDB de lADN initient effectivement la recombinaison miotique (Sun, Treco et al. 1989), et ensuite par des travaux de cartographie haute rsolution des CDB lchelle de chromosomes entiers, en premier lieu chez S. cerevisiae (Baudat et Nicolas 1997; Gerton, DeRisi et al. 2000; Blitzblau, Bell et al. 2007; Buhler, Borde et al. 2007), puis S. pombe (Cromie, Hyppa et al. 2007). Par ailleurs, il sest avr trs tt que la recombinaison miotique aux sites dinitiation produisait systmatiquement des conversions gniques, accompagns dun change des marqueurs flanquants (i.e. dun CO) dans environ 50 % des cas (chez S. cerevisiae) (Hurst, Fogel et al. 1972). Enfin, de fortes variations de la frquence de recombinaison taient observes dune rgion lautre, qui paraissaient dues des variations de la frquence dinitiation aux diffrents sites (Nicolas 1979). Ces sites dinitiation ont t dnomms points chauds (PC) quelques annes plus tard (Cao, Alani et al. 1990; Symington, Brown et al. 1991; Rocco, de Massy et al. 1992; Lichten et Goldman 1995). Au milieu des annes 1990, il semblait donc tabli que la recombinaison miotique tait initie par des CDB formes des frquences variables au niveau de PC localiss en amont des gnes, et produisait la fois des CO et des conversions gniques conformment au modle de DSBR de Szostak. Outre le rle dterminant que le concept dinitiation localise a jou dans llaboration des modles successifs de la recombinaison homologue, les PC ont t par la suite utiliss comme des outils pour ltude des mcanismes sous-jacents, notamment par des approches danalyse physique des intermdiaires de recombinaison. Une question assez intrigante demeurait alors pose: quelle est la nature du dterminisme de la localisation des PC ? En effet chez S. cerevisiae, les analyses de distribution de CDB lchelle dun chromosome (Baudat et Nicolas 1997), ou du gnome entier (Gerton, DeRisi et al. 2000; Buhler, Borde et al. 2007), montrent que la majorit dentre elles est situe en amont des gnes, dans des rgions contenant un promoteur transcriptionnel, mais que ces rgions ne contiennent pas systmatiquement de PC. Par ailleurs les tentatives didentification dun motif de squence conserv, et toujours ncessaire linitiation de la recombinaison, ont toutes choues. Quelques tudes ont dmontr que lactivit de certains PC seulement tait conditionne la prsence de sites de fixation de facteurs de transcription, mais pas la transcription elle-mme (White, Dominska et al. 1993; Mieczkowski, Dominska et al. 2006). Dautres PC semblaient plutt 42

localiss des endroits o la chromatine est dpourvue de nuclosomes (Ohta, Shibata et al. 1994; Kirkpatrick, Wang et al. 1999; Ohta, Wu et al. 1999). En outre la frquence de CDB miotiques semble positivement corrle la proportion locale (par intervalles de 5 kb) de nuclotides G et C dans lADN chromosomique (Gerton, DeRisi et al. 2000). Ces observations diverses ont conduit Petes et al. llaboration dune typologie des PC de recombinaison miotiques chez S. cerevisiae, les trois cas de figure dcrits ci-dessus correspondant respectivement aux PC de type alpha, beta et gamma (Petes 2001). Selon les mmes auteurs, dans les trois situations nanmoins, des modifications dhistones spcifiques pourraient tre impliques dans le recrutement, par un mcanisme inconnu, des protines responsables de la formation des CDB miotiques. Cette hypothse est substantie, plus ou moins directement, par une srie de donnes acquises chez S. cerevisiae, S. pombe et C. elegans (dtailles dans (Borde, Robine et al. 2009)). Effectivement, cette dernire tude, publie en 2009, dmontre que le niveau de CDB miotiques est globalement et localement trs significativement corrl la trimthylation de la lysine 4 de lhistone H3 (H3K4). Ce travail suscite aussi des questions : Certains sites de CDB semblent ne pas dpendre de la prsence de H3K4me3. Il semble donc que dautres modes de dtermination de la localisation des CDB puissent oprer paralllement ou concurremment. La trimthylation de H3K4 marque galement les promoteurs des gnes transcrits durant la miose, mais la relation causale entre transcription et formation des CDB miotiques parat assez obscure. Sur un plan mcanistique, le lien entre la trimthylation de H3K4 et le recrutement des protines impliques dans la formation des CDB demeure inconnu. En 2008 deux articles sont parus (Chen, Tsubouchi et al. 2008; Mancera, Bourgon et al. 2008), qui dcrivent la distribution fine pour lensemble du gnome de S. cerevisiae des vnements de CO et de NCO, indpendamment dans les quatre produits dune srie de miose, la fois dans le type sauvage et plusieurs mutants de gnes impliqus dans le synapsis et/ou la recombinaison miotique. Pour la plupart, les conclusions de ces tudes confirment, prcisent ou tendent ce qui avait t observ prcdemment, notamment dans les analyses de distribution des CDB miotiques (Baudat et Nicolas 1997; Gerton, DeRisi et al. 2000; Buhler, Borde et al. 2007). Lun des rsultats les plus originaux concerne la variation des frquences relatives de CO et de NCO : dans 1,5 % des sites dinitiation la proportion scarte significativement de la moyenne (2 CO pour 1 NCO). Cet cart peut tre est important puisque parfois seul un type dvnement est dtect (les rapports CO/NCO extrmes sont respectivement 14/0 et 0/7) (Mancera, Bourgon et

43

al. 2008). Curieusement, il apparat que la balance CO/NCO est trs abaisse dans les 20 kb adjacents aux tlomres (Chen, Tsubouchi et al. 2008).

1.4.2 Les points chauds de recombinaison miotique chez les animaux et les plantes
1.4.2.1 Lidentification des premiers points chauds de mammifres
Les progrs des techniques de gntique molculaire survenus dans le annes 1970 ont entran lessor des tudes de cartographie physique des gnomes chez les eucaryotes suprieurs (dots dun gros gnome), jusqu lchelle de la squence nuclotidique, et la multiplication des approches de marche chromosomique des fins de clonage positionnel. Ces travaux ont incidemment permis de dcouvrir un fait attendu. Dans certaines zones trs troites (quelques kilobases) du gnome, la frquence de CO miotiques est beaucoup plus leve quen moyenne, et que dans les rgions adjacentes. Par exemple un fragment de 3,5 kb entre les gnes A3 et A2 du complexe majeur dhistocompatibilit (CMH) de la souris contient un CO dans 0,6 % des mioses, ce qui reprsente une frquence (rapporte la taille physique) 300 fois suprieure la moyenne du gnome (Uematsu, Kiefer et al. 1986). Ces rgions dnommes points chauds de recombinaison (avant que ce terme ne soit appliqu par extension aux sites de CDB de S. cerevisiae) ont t identifies initialement dans la rgion du CMH de la souris parce quelle tait activement tudie via des approches de gntique classique (Steinmetz, Uematsu et al. 1987; Shiroishi, Sagai et al. 1993). Par la suite, des PC ont t caractriss galement chez lhomme par des approches danalyse de pedigrees ou dtude du dsquilibre de liaison (DL), et chez le mas en mettant profit la puissance des cribles phnotypiques sur la morphologie et la couleur du grain (Dooner 1986; Xu, Hsia et al. 1995). Le terme de point chaud tait alors frquemment (et est encore aujourdhui parfois) utilis en rfrence des rgions du gnome plus chaudes quen moyenne, sans considration ni de leur taille, ni de la distribution fine des CO quelle contiennent. Cette dfinition oprationnelle plutt floue contraste avec le modle de PC dfini chez S. cerevisiae dune manire beaucoup plus prcise et formelle. Cela peut certainement tre imput notamment la difficult de disposer dun nombre suffisamment lev dvnements. En effet, au niveau de la plupart des PC, la densit de CO est effectivement trs accrue comparativement la moyenne du gnome, mais leur frquence totale est nanmoins gnralement trs faible, et dpasse rarement 1 % (0,3 % au PC SHOX chez H. sapiens (May, Shone et al. 2002)). Cela implique que, par des approches gntiques classiques,

44

des populations de plusieurs milliers dindividus sont requises pour disposer de quelques dizaines dvnements aux points les plus chauds . A cet gard, le dveloppement de la PCR a beaucoup contribu amliorer la qualit des tudes de PC, en premier lieu chez les mammifres. Tout dabord, la mise au point de mthodes de cartographie gntique bases sur la PCR, partir de spermatozodes isols (Boehnke, Arnheim et al. 1989), a permis dobtenir par exemple une carte dtaille du CMH humain (Hubert, MacDonald et al. 1994; Cullen, Perfetto et al. 2002) et dautres rgions de grande taille (Lien, Szyda et al. 2000; Greenawalt, Cui et al. 2006). Toutefois ces tudes, bien que permettant didentifier des rgions susceptibles de contenir des PC, manquaient de la puissance requise pour les tudier finement.

1.4.2.2 Les points chauds de mammifres caractriss par typage de sperme


Cest une innovation technologique supplmentaire qui va ouvrir la voie la caractrisation trs prcise de PC dans les gnomes danimaux et de plantes. Elle consiste amplifier spcifiquement les molcules recombines (a priori essentiellement les CO) dans des pools dADN gnomique extrait de cellules gamtiques/gamtophytiques produites par des hybrides (voir les dtails dans la section 4). Cette technique de typage de sperme , mise au point en 1998 par A. Jeffreys (Jeffreys, Murray et al. 1998) a permis depuis dtudier en dtails (gnralement) 36 PC chez lhumain et 13 chez la souris. Le total peut sembler modeste, au regard de lintrt que reprsente a priori ce type dapproche, mais il sexplique par deux raisons. Tout dabord, le typage de sperme ne sapplique qu ltude de PC uniques ou trs voisins prcdemment identifis, ou au moins suspects. Il faut donc avoir au pralable men une analyse gntique plus classique, permettant de les localiser une chelle rduite (quelques kb), avec toute la difficult que cela suppose (voir plus haut). Actuellement nanmoins, limportance de cette limitation va dcroissant du fait que les approches de caractrisation gnomique (squenage, hybridation de puces) ont tendance se banaliser. En fait, 19 des 36 PC caractriss chez lhomme ont t initialement identifis par lanalyse haute rsolution du dsquilibre de liaison lchelle du gnome entier (Webb, Berg et al. 2008; Jeffreys et Neumann 2009). En second lieu, le succs de lanalyse par typage de sperme repose sur lexistence de polymorphismes permettant (1) damplifier spcifiquement les molcules recombines par PCR et (2) de localiser les transitions gnotypiques avec une rsolution suffisante. Si lon fait abstraction de ces limitations intrinsques, le typage de sperme est une mthode extrmement puissante danalyse des CO. En thorie elle permet de caractriser un nombre virtuellement infini dvnements survenant des frquences ventuellement trs faibles.

45

Nombre de Largeur 95% Point chaud CO analyss CO (kb) DNA1 69 1,9 DNA2 237 1,3 DNA3 661 1,2 DMB1 36 1,8 DMB2 358 1,2 TAP2 141 1 MS32 250 1,5 SHOX 527 1,9-2,5 NID1 1345 1,5 NID2a 302 1,4 NID2b 107 1,1 NID3 1094 2 MSTM1a 179 1,6 MSTM1b 374 2,1 MSTM2 46 1,3 DPA1 130 1,7 Beta globine 283 1,2 Localisation promoteur intergnique intergnique intron/exon intragnique intron intergnique intron/exon intron intron intron intergnique intergnique intergnique intergnique intergnique intergnique Espce H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens H. sapiens

Frquence des Frquence de CO x 10-5 pic (cM/Mb) 0,5 0,4 3,7 2,5 130 140 3,1 2 28 20 5,8 9 39 40 300 190-370 67 70 8,5 10 3 4 90 70 8,8 9 15 16 0,7 0,9 29 27 150 200

Sources (Jeffreys, Kauppi et al. 2001) (Jeffreys, Kauppi et al. 2001) (Jeffreys, Kauppi et al. 2001) (Jeffreys, Kauppi et al. 2001) (Jeffreys, Kauppi et al. 2001) (Jeffreys, Ritchie et al. 2000) (Jeffreys, Neumann et al. 2005) (May, Shone et al. 2002) (Jeffreys, Neumann et al. 2005) (Jeffreys, Neumann et al. 2005) (Jeffreys, Neumann et al. 2005) (Jeffreys, Neumann et al. 2005) (Jeffreys, Neumann et al. 2005) (Jeffreys, Neumann et al. 2005) (Jeffreys, Neumann et al. 2005) (Kauppi, Stumpf et al. 2005) (Holloway, Lawson et al. 2006)

Table 3 (dbut). Liste des points chauds de crossovers tudis par typage de sperme chez les mammifres.

En pratique cest une technique dont la mise au point et la mise en uvre peuvent savrer laborieuses (section 4). Quand des polymorphismes adquats sont prsents lintrieur mme des PC, le typage de sperme peut aussi tre appliqu la caractrisation dvnements de conversion gnique. Nanmoins, il sagit l galement dune limitation contingente, qui savre rdhibitoire la plupart du temps : les conversions gniques doivent alors tre tudies par une autre approche (section 2.3.2.). Les donnes factuelles publies, issues de lanalyse de PC de mammifres par typage de sperme, sont rcapitules dans la table 3. Plusieurs faits essentiels sont ressortis des tudes de PC par typage de sperme (May, Slingsby et al. 2008): Les PC ne semblent localiss prfrentiellement ni dans des exons, ni dans des promoteurs. La distribution des CO est en rgle gnrale globalement symtrique (autant quil soit possible den juger compte tenu de la densit des polymorphismes, parfois trs inhomogne), et apparemment conforme une distribution normale (Jeffreys, Kauppi et al. 2001) La taille des PC est assez constante, de lordre de 1 2 kb. Le PC M2 de souris (Kauppi, Jasin et al. 2007) constitue une exception notable, mais il se peut quil sagisse dun PC composite. Les PC ne semblent en effet pas rpartis de faon homogne : ils sont parfois groups, au point que les distributions de CO peuvent se chevaucher. Les PC fusionns pourraient en fait assez frquents (voir plus bas, et sections 2.3. et 3.4.). Il y a des points brlants et des points peine tides, entre lesquels la frquence de CO peut varier dun facteur 1000. A certains PC et dans certains croisements, la reprsentation alllique parmi les CO est biaise au centre du PC. Cela reflte un dcalage de la position moyenne des transitions de gnotype dans un sens et dans lautre. Lhypothse la plus communment avance pour expliquer ce fait tient ce que les CO ne sont des vnements rciproques qu distance des CDB initiatrices : proximit immdiate la portion dADN casse et rsecte est rpare par conversion gnique. Si les CDB ne se produisent pas la mme frquence sur les deux haplotypes parentaux, alors la conversion gnique ne les affecte pas galement, ce qui aboutit un biais de reprsentation (Jeffreys et Neumann 2002; Yauk, Bois et al. 2003; Jeffreys, Holloway et al. 2004; Jeffreys et Neumann 2005; Baudat et de Massy 2007; Bois 2007)(figure 21). La variation de lactivit initiatrice des diffrents allles se traduit galement par une variation de la frquence de CO selon les fonds gntiques, soit des individus diffrents chez lhomme, soit des hybrides entre souches diffrentes chez la souris. De mme quil y a des points brlants et

46

Nombre de Largeur 95% Point chaud CO analyss CO (kb) Localisation Espce Sources

Frquence des Frquence de CO x 10-5 pic (cM/Mb)

A, B, C1, C2, D, E, F, G1, G2, H, J1, J2, K, L, M, N, P, Q, R Psmb9 Ebeta HS22 M1 M2 Hlx1 HS14.9 HS18.2 HS23.9 HS44.2 HS48.3 HS61.1 HS61.2 15 - 1000 550 159 217 5 7 2800 5,1 14,7 2,5 21,1 0,9 49,9 55,8 650 140 62 3 <2 ND ND ND ND ND ND ND ND H. sapiens M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus M. musculus intergnique (13) /intron (6) intergnique intron/exon intron intergnique intergenique ND ND ND ND ND ND ND ND

~110 69 124 327 75 57 14 91 122 144 71 87 102 210

1,2 1,9 1,8 1,4 2 2 9 1,5 ND ND ND ND ND ND ND

(Webb, Berg et al. 2008) (Guillon et de Massy 2002) (Yauk, Bois et al. 2003) (Bois 2007) (Kauppi, Jasin et al. 2007) (Kauppi, Jasin et al. 2007) (Ng, Parvanov et al. 2008) (Wu, Getun et al. 2010) (Wu, Getun et al. 2010) (Wu, Getun et al. 2010) (Wu, Getun et al. 2010) (Wu, Getun et al. 2010) (Wu, Getun et al. 2010) (Wu, Getun et al. 2010)

Table 3 (fin). Liste des points chauds de crossovers tudis par typage de sperme chez les mammifres.

des points tides, il y a des allles brlants et des allles froids. Cela a t dcrit notamment pour les PC Psmb9 murin (Baudat et de Massy 2007; Grey, Baudat et al. 2009) et S2 humain (Jeffreys et Neumann 2009). Cette rgulation de lactivit de recombinaison locale exerce par des dterminants locaux est qualifie de contrle en cis. Lactivit dinitiation un PC peut galement dpendre de polymorphismes situs un autre endroit du gnome : il sagit dans ce cas dun contrle exerc en trans (voir plus bas). Il arrive quune partie des CO prsente un profil complexe de transition entre les gnotypes parentaux. Le long de ces molcules, on observe dans la rgion du PC une alternance de segments soit spcifiques de lun ou lautre parent, soit htroduplex. Gnralement, ces molcules complexes sont rares (1 % ou moins), mais leur proportion peut atteindre 4% (Bois 2007), voire 25 % (Kauppi, Jasin et al. 2007) certains PC. Ces structures en mosaque sont vraisemblablement issues dune rparation dficiente ou aberrante des htroduplex qui sont forms au cours du processus de recombinaison homologue, soit par linvasion de brin et la capture de la deuxime extrmit, soit par lextension des extrmits 3 par synthse dADN partir dune matrice homologue, soit par la migration des jonctions de Holliday. Il a t propos que la frquence et la nature (INDEL vs SNP) des polymorphismes prsents dans les PC puissent dterminer ces accidents de rparation des rgions msapparies (Bois 2007) (voir sections 2.3. et 3.4.). La caractrisation dvnements de NCO aux PC de mammifres (ou de plantes) par les mthodes molculaires prsente gnralement une difficult technique telle que les donnes publies sont rares (table 4). Ces donnes concordent avec ce qui a t observ chez S. cerevisiae (section 1.4.1.) : la balance entre les frquences de CO et de NCO fluctue beaucoup, environ entre 3 et moins de 1/12, alors que la moyenne gnomique attendue chez lhumain est infrieure 1/6 (47 foci MLH1 pour environ 300 foci RAD51). Par contre la taille des tracts de conversion gnique associs aux NCO est nettement infrieure (sans doute moins de 540 pb) ce qui est observ chez S. cerevisiae (1.8 kb en moyenne) (Chen, Tsubouchi et al. 2008; Mancera, Bourgon et al. 2008).

1.4.2.3 Les points chauds de plantes


Jusqu prsent, la distribution des vnements de recombinaison petite chelle a t trs peu tudie chez les plantes, comparitivement aux animaux. Seules deux rgions ont t dcrites en dtails, chez le mas : la gne bronze (bz) et la rgion anthocyaninless 1-shrunken kernel 2 (a1-sh2). Lexistence dune srie dallles du gne bz a permis disoler, grce un crible phnotypique portant sur la couleur du grain, un grand nombre dvnements de recombinaison intragniques,

47

Point chaud Beta globine DNA3 DMB2 NID1 SHOX Psmb9 Hlx1 Esrrg1

Frquence de Frquence de CO x 10-5 NCO x 10-5 290 <25 110 300 5 4 50 13 370 90 550 270 2800 1600 2000 2500

Ratio CO:NCO <1:12 2,7:1 01:01,3 <1:4 01:03,3 02:01 1,8:1 1:1;2

Taille des tracts (pb) Espce 60-300 H. sapiens 55-280 H. sapiens H. sapiens <1200 H. sapiens H. sapiens <540 M. musculus ND M. musculus ND M. musculus

Sources (Holloway, Lawson et al. 2006) (Jeffreys et May 2004) (Jeffreys et May 2004) (Jeffreys, Neumann et al. 2005) (Jeffreys et May 2004) (Guillon et de Massy 2002) (Parvanov, Ng et al. 2009) (Parvanov, Ng et al. 2009)

Table 4. Liste des points chauds de CO et NCO tudis par typage de sperme chez les mammifres.

des CO aussi bien que des NCO (Dooner et Martinez-Ferez 1997; Dooner 2002). Ces rsultats ont t obtenus dans des contextes hybrides varis, et la localisation prcise des vnements de recombinaison na pas t dtermine distinctement pour les CO et les NCO. De plus la frquence et la position des vnements de recombinaison dans les rgions flanquant le gne bz sont galement inconnues. Tous les lments ne sont donc pas runis, qui permettraient daffirmer que le locus bz est un PC au sens dfini chez les champignons et les mammifres. Il se peut cependant que le gne bz prsente un gradient de polarit de recombinaison, ce qui est caractristique dun PC, mais cette question a fait lobjet dun dbat qui nest toujours pas tranch (Dooner 1998; Thijs et Heyting 1998). La rgion a1-sh2 a galement t caractrise laide dun crible phnotypique, portant sur la forme et la couleur du grain, mais sans a priori sur la localisation intra ou extra-gnique des vnements de recombinaison. La distribution fine des CO, tablie partir de 176 vnements dans un des contextes hybrides tudis, fait apparatre trois pics distincts dans une rgion de 10 kb entre les gnes a1 et yz1, et un pic isol distant de 35 kb (Yao et Schnable 2005)(figure 22). Ce tableau est conforme au modle de PC valid chez les champignons et les animaux. En outre 8 vnements de NCO putatifs localiss totalement ou en partie dans le gne a1 ont t identifis (Yandeau-Nelson, Zhou et al. 2005). La longueur de la majorit (7/8) des tracts de conversion gnique associs dpasse 500 pb, ce qui semble les distinguer des NCO caractriss dans les PC de mammifres, dont les tracts de conversion gnique dpassent rarement 500 pb (section 1.4.2.2.). En labsence de marqueurs phnotypiques prsentant une liaison gntique quantifiable tout en tant physiquement trs proches, comme dans le cas de a1 et sh2, lidentification de PC repose sur des stratgies de marche chromosomique utilisant des marqueurs molculaires anonymes. La mise en uvre de ces approches est dautant plus longue et laborieuse que les chromosomes du gnome tudi sont physiquement grands. Nanmoins, les progrs des technologies de cartographie physique et de typage gntique permettent denvisager ltude systmatique de la recombinaison miotique petite chelle dans des espces petit gnome, comme A. thaliana (Drouaud, Camilleri et al. 2006), ou gnome volumineux, comme le bl (Saintenac, Falque et al. 2009). Il semble donc probable que le nombre de PC caractriss en dtails chez les plantes saccroisse beaucoup dans un avenir assez proche.

48

60 taux de crossovers (cM/Mb) 50 40 30 20 10


1 4

37 24

60

8 17 1

0 0 2 4 6 coordonne physique (kb) 8 10

a1

yz1

Figure 22. Distribution des crossovers dans la rgion a1-yz1 chez le mas. 161 crossovers ont t identifis et localiss dans la descendance dun hybride entre deux lignes quasi isogniques de Zea mays spp. mays portant des haplotypes a1 sh2 diffrents. Le nombre de CO est indiqu au dessus de chaque intervalle. Daprs (Yao et Schnable 2005).

1.4.2.4 Le projet HapMap et les points chauds historiques chez lhomme


Lune des limitations essentielles ltude des PC de recombinaison dans les gnomes de grande taille concerne leur localisation prliminaire. Un PC ne peut tre analys en dtails que sil a t identifi et positionn dans un intervalle physique relativement petit. Les mthodes traditionnelles de cartographie gntique peuvent tre utilises dans des espces modles comme la souris, mais chez lhumain lanalyse de pedigrees est comparativement beaucoup moins puissante. Le typage de spermatozodes (voir plus haut) donne accs un nombre potentiellement trs lev de recombinants, mais cette stratgie a finalement peu t utilise, probablement en raison de limitations techniques. Chez lhumain, cest ltude du DL qui sest avre tre lapproche la plus puissante pour dtecter les rgions chaudes , donc susceptibles de contenir des PC de recombinaison (Chakravarti, Buetow et al. 1984; Chakravarti, Elbein et al. 1986; Yip, Lovegrove et al. 1999; Jeffreys, Kauppi et al. 2001; May, Shone et al. 2002; Kauppi, Sajantila et al. 2003; Rana, Ebenezer et al. 2004). Nanmoins lchelle de populations, les variations du DL rsultent non seulement de la distribution inhomogne des CO le long des chromosomes, mais aussi de phnomnes multiples lis la slection et lhistoire dmographique (drive gntique, variations deffectif et goulots dtranglement, migrations et mlanges de populations). Ainsi, dans la rgion DPB1 du CMH, par exemple, on observe bien une rupture du DL, mais le typage de sperme rvle quen fait les CO y sont trs peu frquents et distribus de faon homogne, c'est--dire quil ny a pas de PC (Kauppi, Stumpf et al. 2005). A linverse, il arrive que des PC soient dtects dans des rgions prsentant une faible diminution du DL. Cest le cas notamment prs du gne NID3 (Jeffreys, Neumann et al. 2005). Ces exemples suggrent que (i) des PC apparus rcemment pourraient ne pas avoir eu le temps de marquer de leur empreinte la structure haplotypique des populations et que (ii) des ruptures de LD pourraient avoir t causes par des PC actuellement disparus. Cette question de la correspondance entre PC - actuels ou fossiles - et structure haplotypique a suscit le dbat partir du moment o des donnes de gnotypage couvrant lensemble du gnome humain, avec une densit de marqueurs lev, et chez une srie dindividus chantillonns dans des populations gntiquement diffrenties, ont t produites par le consortium international HapMap (HapMap 2005; HapMap 2007). Face aux sceptiques arguant que les PC, sils existent, ne peuvent expliquer seuls la structure haplotypique du gnome (Tishkoff et Verrelli 2003), et que les limites entre blocs haplotypiques ne correspondent gnralement pas aux PC (Wall et Pritchard 2003), les optimistes ont rtorqu que le DL est de

49

fait corrl ngativement au taux de recombinaison actuel (Greenwood, Rana et al. 2004; Greenawalt, Cui et al. 2006) et dvelopp des mthodes danalyse fine permettant selon eux de dtecter les PC actifs au cours de lhistoire des populations (Li et Stephens 2003; Fearnhead, Harding et al. 2004; McVean, Myers et al. 2004). Ainsi, en 2005 une tude est publie qui dcrit lidentification de 25000 points chauds historiques (PCH) pour lensemble du gnome humain (Myers, Bottolo et al. 2005) (sujet revu par (Myers et McCarroll 2006)). Les auteurs estiment que cela reprsente environ 50 % des PC existant actuellement, et en dduisent quil existe un PC actif tous les 50 kb, ce qui correspond la densit observe dans le CMH de classe II (Jeffreys, Kauppi et al. 2001). Par ailleurs, des analyses de corrlation sont menes qui montrent notamment que les rtrotransposons de type THE1A et THE1B sont fortement surreprsents dans les PCH, et que ces rtrolments diffrent de leurs collgues absents des PCH par la prsence trs frquente dun motif de squence particulier CCTCCCT. Plus rcemment, une seconde tude ralise par la mme quipe (Myers, Freeman et al. 2008), partir des donnes supplmentaires produits par le projet HapMap (HapMap 2007), a permis daffiner les rsultats de lanalyse prcdente : le motif dgnr CCTCCCTNNCCAC dans le contexte dun lment THE1A est associ un PCH dans 73 % des cas, et seulement 10 % des cas hors de ce contexte. Les auteurs remarquent en guise de conclusion que ce motif pourrait tre li par une protine doigts de zinc. Au passage, il est intressant de noter que le motif CCTCCCT est prsent au centre du PC DNA2 dans un allle chaud , mais pas dans un allle froid qui prsente une substitution rompant le consensus. Ces rsultats trs originaux ont sembl dautant plus convaincants que des tudes de validation grande chelle ont ensuit t publies, qui montrent que la plupart des PCH sont des PC actuels, et rciproquement, bien quapparemment il existe parfois des diffrences importantes dans lusage des PC actuels entre les individus (Tiemann-Boege, Calabrese et al. 2006; Coop, Wen et al. 2008).

1.4.2.5 Vers llucidation du dterminisme de localisation des points chauds de mammifres


Lhistoire du mystrieux motif CCTCCCTNNCCAC ne sest pas arrte en 2008. En 2010, lquipe dOxford lorigine de cette dcouverte rvle que le gne Prdm9 code chez lhomme la protine doigts de zinc la plus susceptible de lier le motif CCTCCCTNNCCAC (Myers, Bowden et al. 2010). Mais pas chez le chimpanz, le macaque, lorang-outan, la souris, le rat ou llphant En effet ce gne volue trs rapidement. Or, il se trouve que ni les PC ni les PCH humains ne semblent conservs chez le chimpanz (Ptak, Roeder et al. 2004; Ptak, Hinds et al. 50

2005; Winckler, Myers et al. 2005), en dpit de lidentit nuclotidique trs leve (99 %) entre les rgions colinaires de leurs gnomes. Il tait connu depuis 2005 que chez la souris, la protine PRDM9 (initialement dnomme MEISETZ) est implique dans la trimthylation de la lysine 4 de lhistone H3, et que sa prsence est en outre ncessaire la formation des CDB miotiques (Hayashi, Yoshida et al. 2005). Quelques annes plus tard, la relation entre ces deux activits a t dmontre : chez la souris, la chromatine du points chaud Psmb9 dans des spermatocytes au stade pachytne contient spcifiquement (c'est--dire pas dans des cellules de foie, et pas dans les rgions froides immdiatement adjacentes au PC) des histones modifies. Certaines modifications caractrisent un allle froid de Psmb9, tandis que dautres, dont H3K4Me3, sont associes un allle chaud. H3K4me3 est prsente en labsence de CDB miotiques (i.e. dans un mutant Spo11 -/-), et donc ventuellement ncessaire leur formation. Par ailleurs, lactivit des allles chauds et la prsence des modifications dhistones caractristiques sont dpendantes dun haplotype donn dans une rgion distante dau moins 10 Mb (Buard, Barthes et al. 2009). Une seconde tude produite par la mme quipe montre que le locus responsable de ce contrle en trans de lactivit de Psmb9 module en fait, non pas le niveau global, mais la distribution des vnements de recombinaison miotiques, c'est--dire la fois les CDB et les CO de type I, lchelle de tout le chromosome 17 (contenant Psmb9) et probablement du gnome entier (Grey, Baudat et al. 2009). Ce locus baptis Dsbc1 (double-strand break control 1) est par la suite localis plus finement dans un intervalle de 4,6 Mb du chromosome 17, qui savre contenir . Prdm9 (Baudat, Buard et al. 2010). Chez la souris, les allles de Prdm9 prsents dans les haplotypes activateur ou nonactivateur de Psmb9 sont effectivement diffrents, et lanalyse de squence prdit que les variants protiques correspondants sont dots chacun dune spcificit de liaison lADN distincte. Chez lhumain, dans la population hutterite, le gne Prdm9 prsente une variabilit alllique importante, qui semble associe la variation dusage des PCH voque la fin de la section 1.4.2.3., et qui dtermine in vitro la modulation de la spcificit de liaison de la protine au motif CCTCCCTNNCCAC (Baudat, Buard et al. 2010). Ces travaux dmontrent le rle de PRDM9 dans la rgulation de lactivit, et corollairement de la distribution, dune partie des PC du gnome des mammifres. Lhypothse selon laquelle la distribution des vnements de recombinaison miotique serait rgule par des protines voluant rapidement, dont PRDM9 serait un archtype, fournit une solution trs commode au paradoxe du point chaud . Cette hypothse prdit que puisque les allles chauds , initiateurs de la recombinaison miotique sont convertis en allles froids dans les htrozygotes par le seul mcanisme de la recombinaison homologue, leur frquence doit 51

peu peu baisser dans les populations, si bien que les PC refroidissant avec le temps sont censs disparatre finalement (Boulton, Myers et al. 1997). On peut arguer que ce mcanisme nest actif que dans les espces prsentant un niveau de polymorphisme intra-populationnel lev, ce qui nest pas le cas quand lautofcondation est le mode de reproduction sexue majoritaire, ou quand le sexe est utilis occasionnellement. Nanmoins, les tudes par typage de sperme dmontrent que les PC peuvent effectivement disparatre chez lhomme, au moins dune gnration la suivante (voir section 1.4.2.2.). Le polymorphisme de Prdm9 observ dans les populations humaines suggre prcisment que les PC peuvent bien disparatre petit petit, puisquil existe un mcanisme permettant den crer dautres rapidement. Des questions demeurent en suspens : de quelle manire PRDM9 cible telle lactivit de cassure double-brin catalyse par SPO11 ? Quel est le rle propre de la mthylation de H3K4 ? Comment les autres protines impliques dans la formation des CDB sont-elles impliques ? Il existe dsormais un faisceau darguments pointant le rle des modifications dhistones dans le dterminisme de la formation des CDB, aussi bien chez S. cerevisiae que chez les mammifres. Il est donc tentant dimaginer que ce mcanisme est universel, y compris chez lez plantes, dans lesquelles la distribution des vnements de recombinaison semble obir un modle de points chauds similaire celui des mammifres.

52

2 Rsultats
Les rsultats obtenus au cours de la prparation de cette thse seront prsents dans le cadre de trois articles. Le premier est une revue dite en 2007, crite en collaboration avec Raphal Mercier, Christine Mzard et Julien Vignard. Elle traite de la distribution chromosomique des CO, de leffet de linterfrence et des voies de formation de CO, interfrents et non-interfrents. Ma contribution porte essentiellement sur la description des mthodes de mesure de linterfrence et sur la prsentation des modles biologiques sous-jacents. Le deuxime article, galement paru en 2007, concerne la distribution des CO le long du chromosome 4 dA. thaliana, dans la miose des deux sexes. Ce travail sinscrit dans la continuit dun travail initi en 2004, qui avait dj donn lieu une publication en 2006. Cette dernire tude, bien que ne faisant pas partie de la thse, est nanmoins annexe au manuscrit (section 6), afin de permettre une meilleure apprhension du contexte des travaux qui ont suivi. Le troisime article est sur le point dtre soumis pour publication au journal Science . Il prsente ltude dtaille de deux points chauds de recombinaison miotiques, dnomms 14a et 130x , localiss sur le chromosome 4 dArabidopsis thaliana. Ds le dbut de ma thse, jai entrepris de mettre au point la mthode de pollen typing , ce qui ma ensuite permis de caractriser en dtails le point chaud 130x. Le point chaud 14a est tudi par Hossein Khademian dans le cadre de la prparation de sa propre thse de Doctorat dUniversit.

2.1 Les voies des crossovers : la parole est aux plantes

53

Review

TRENDS in Genetics

Vol.23 No.2

The road to crossovers: plants have their say


zard, Julien Vignard, Jan Drouaud and Raphae Christine Me l Mercier
Station de Ge ne tique et dAme lioration des Plantes, Institut Jean Pierre Bourgin, INRA, 78026 Versailles cedex, France

Crossovers involve the reciprocal exchange of large fragments of genetic material between homologous chromosomes during meiosis. In this way, crossovers are the basis of genetics. Remarkably, the number and distribution of crossovers on chromosomes are closely controlled. Data from various model organisms (notably Saccharomyces cerevisiae) show that the distribution of crossovers results from a series of tightly regulated events involving the formation and repair of doublestrand breaks and interference. Recent advances in genetic and cytological tools, particularly for studying Arabidopsis thaliana, have enabled crossover control in plants to be studied in more detail. In this article, we discuss the contribution of plant studies to meiosis research, particularly to our understanding of crossover control and interference, and we evaluate models of interference. Introduction The crossover (see Glossary) and its cytological signature, the chiasma, are major features of genetics. Our knowledge of the molecular mechanisms of crossover formation has increased considerably in the past decade, owing to studies of fungi [1,2]. Nevertheless, little is known about controlling the number of crossovers and their distribution along chromosomes, except for the remarkable observation that crossovers and/or chiasma are not randomly distributed [3]. However, it has been observed that when this control is affected, the missegregation of chromosomes markedly increases, resulting in aneuploid gametes in organisms as diverse as yeast, nematodes, mammals and plants [4]. The current view is that the distribution of crossovers results from three constraints. First, each pair of chromosomes, regardless of size, has at least one chiasma (known as the obligatory chiasma), which is essential for the segregation of homologous chromosomes at the rst meiotic division. Second, crossovers are not independent of each other: the occurrence of one crossover inhibits the occurrence of another such event in a distance-dependent manner, resulting in crossovers being spaced more evenly along the chromosomes than would be expected if they occurred independently. This phenomenon is known as positive interference and is referred to as interference throughout this article [3]. Third, separately from the inuence of interference, the densities of crossovers along chromo zard, C. (mezard@versailles.inra.fr); Corresponding authors: Me Mercier, R. (rmercier@versailles.inra.fr). Available online 8 January 2007.
www.sciencedirect.com

somes vary greatly, so there is a nonhomogeneous distribution of crossovers. Plants have always been at the forefront of the study of heredity (Box 1). Now, plant models can be used to study meiosis and recombination. This is because they not only enable use of the rare combination of both genetic and cytological approaches but also, owing mainly to technological progress with Arabidopsis thaliana, a full range of molecular approaches can be used. The data obtained in plants contribute to our general understanding of meiosis, Glossary
Chiasma: the cytological signature of a crossover. Chiasmata are observed as connections between the homologous chromosomes in bivalents during meiosis, from diakinesis to metaphase I. Crossovers: one of the products of meiotic DSB repair. The repair process, through breaking and rejoining DNA molecules, results in the reciprocal exchange of large fragments of genetic material (i.e. the exchange of homologous regions). Therefore, the genetic outcome of crossovers is the reassociation of genetic markers located on both sides of the crossover point. Also called crossing-overs. Gene conversion: the nonreciprocal exchange of small fragments of genetic material (i.e. homologous regions), resulting in a non-mendelian segregation of genetic markers. Gene conversions are formed by the repair of a short region around a DSB site by homologous recombination, using the chromatid of the homologous partner chromosome as a template. They can be single or associated with a crossover. Leptotene: the first of the four substages of prophase in the first meiotic division (i.e. prophase I). During this substage, individual chromosomes condense into long strands. Initiation of recombination (i.e. DSBs) occurs at this stage. Noncrossovers: the subset of meiotic DSB-repair events that is not associated with crossovers. Noncrossovers can be genetically detectable, in which case they are called single gene-conversion events. Whether noncrossovers are genetically detectable depends on the presence or absence of a polymorphism in the conversion tract (i.e. the fragment of DNA that has been exchanged in the process). Pachytene: the third of the four substages of prophase in the first meiotic division (i.e. prophase I). This substage is characterized by completion of the polymerization of the synaptonemal complex along homologous chromosomes. Recombination is completed at this stage. Recombination nodules: electron-dense structures located on chromosomes. These are observed in electron micrographs of meiotic chromosomes during prophase I. During zygotene, the early nodules (ENs) have a variable shape: for example, round and small in tomato (Solanum lycopersicum), and ellipsoidal and spherical in Allium species. These nodules disappear early in pachytene. The nodules that remain at pachytene are called late nodules (LNs), and they are larger and are usually ellipsoidal. Synaptonemal complexes: structures that are specific to meiosis. By electron microscopy, a synaptonemal complex appears as a ladder-like structure that forms at zygotene, and formation of this complex is completed at pachytene. The two uprights of the ladder consist of each axial element (the continuous protein structures formed along each chromosome at leptotene) of the two homologous chromosomes (at this stage called lateral elements), and the rungs consist of a dimer of the protein Zip1. The function of synaptonemal complexes has not been determined. Zygotene: the second of the four substages of prophase in the first meiotic division (i.e. prophase I). This substage is characterized by formation of the synaptonemal complex, which progressively links closely the pairs of homologous chromosomes. Recombination progresses during this stage.

0168-9525/$ see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.tig.2006.12.007

54

92

Review

TRENDS in Genetics

Vol.23 No.2

Box 1. Genetics in plants: from Pisum sativum to Arabidopsis thaliana


Since Mendel established the law of segregation of independent characteristics using the pea (Pisum sativum), plants have been at the heart of the history of genetics. Mendels work was ignored for three decades, until his laws were rediscovered independently through studies of a large range of plants, including Trifolium pratense (red clover), Silene latifolia subsp. alba (white campion), Papaver somniferum (opium poppy), Zea mays (maize) and P. sativum [69,70]. The first partial linkage was found in Lathyrus odoratus (sweet pea) [71]. Although the data were not initially interpreted as a meiotic reassociation of characteristics, this was the first report of a genetic crossover. The first demonstration that crossovers are associated with physical recombination of chromosomes was in Z. mays [72], and the first evidence for chiasmata interference, predicted from crossover interference, was obtained in Vicia faba (fava bean) [70,73]. Plants have been fundamental in numerous other genetic discoveries. Mobile DNA elements were first reported in Z. mays [74]. RNA-mediated silencing was first discovered in transgenic Petunia hybrida, and plants still have a central role in deciphering the underlying mechanisms of gene silencing [75]. One of the current favorite models of plant geneticists is Arabidopsis thaliana because of its suitability for combined molecular genetic and cytological analyses.

particularly owing to the comparisons that can be made with data obtained in other model organisms. Here, we review recent ndings on crossovers in plants, particularly those concerning key genes in Arabidopsis, and we relate these ndings to the numerous genetic and cytological data from other species, including yeast, on the distribution of crossovers and the role of interference. Different manifestations of recombination: from yeast to plants Recombination was rst dened as the generation of a new combination of genes [5]. Now, this denition can be extended to a reassortment of markers. In meiosis, this can arise either by crossover or by single gene conversion. Recently, the term noncrossover has been introduced to designate DNA double-strand break (DSB)-repair events that are not associated with crossovers. Noncrossovers include single gene-conversion events and events that are not genetically detectable. Crossover and noncrossover events are products of the repair of programmed DSBs [6]. It is also probable that some repair events occur on the sister chromatid [sister chromatid exchange (i.e. SCE)]; however, these events are difcult to analyze, so their frequency is not easy to estimate [7]. In budding yeast (Saccharomyces cerevisiae), meiotic DSBs are caused by a set of 11 proteins, including Spo11, which directly generates the DSBs [6] (Figure 1). Only four of these proteins (Spo11, Mre11, Rad50 and Ski8) are conserved in Arabidopsis. However, there are three Spo11 homologs in Arabidopsis. The Arabidopsis proteins AtSPO11-1 and AtSPO11-2 are required to initiate meiotic recombination [8,9]. Mre11 and Rad50 are involved in both the formation and the repair of DSBs in yeast, but their orthologs are not required for DSB formation in Arabidopsis [10,11]. It is also clear that the homolog of Ski8 is not required for meiosis in plants [12]. Although the mechanism of initiation of DSB formation is conserved across all species, there are clear differences in the control of this process.
www.sciencedirect.com

Numerous data on crossover distribution in plants are available, largely because crossovers are the basis of genetic maps. By contrast, information about noncrossover distribution is poor, because these events are difcult to visualize by classical genetic analyses and have no (or little) impact on genetic maps. Crossovers can also be studied by cytology, visualized as chiasmata or as late recombination nodules (LNs). Chiasmata reect the same molecular event as crossovers [3] and are powerful and easy markers for counting the number of crossovers per cell and per chromosome. However, counting in this way could result in a slight underestimation, because, even in plants with large genomes, chiasmata that are close together cannot be resolved unambiguously [13]. Individual chiasmata can be visualized only in a few organisms, such as grasshoppers [3]. LNs lie on the central region of synaptonemal complexes during pachytene and closely reect the sites of crossovers, providing a much better resolution than that provided by chiasmata [14,15]. However, the technique for efcient visualization of LNs is applicable to only a few species. At an earlier substage in prophase I (zygotene), there is a much larger population of nodules, called early recombination nodules (ENs). Their distribution, their protein content and the timing of their appearance indicate that ENs might mark the sites of recombination intermediates [14]. The recombinases Rad51 and Dmc1 are involved in meiotic DSB repair and colocalize with ENs [16]. These two proteins are not observed in all ENs at a given time, but this is probably because ENs have progressed through the process of DSB repair to different extents. The genes ATRAD51 and ATDMC1 have been functionally characterized in Arabidopsis. As expected, the atrad51 mutant shows strong meiotic fragmentation of its chromosomes [17]. Surprisingly, in Arabidopsis atdmc1 mutants, DSBs seem to be repaired using the sister chromatid, although they are not repaired in similar yeast mutants [18,19]. Another interesting difference is that recombination defects result in apoptosis in mammals, whereas, in Arabidopsis, meiosis progresses regardless of the defect, allowing access to more information about the mutant phenotype. Crossover frequency varies along chromosomes In all eukaryotes that have been studied, including plants, the distribution of crossovers or LNs along chromosomes is not homogeneous [15,20], on either a megabase or a kilobase scale (Figure 2). That is, the local probability of a crossover varies among different chromosome intervals. This nonhomogeneity is the basis of the denition of hot and cold regions (which have signicantly high and low crossover frequencies, respectively). A general rule is that centromeres are cold regions [21]. Plants have diverse chromosome sizes and structures, and, in each type of plant, these hot and cold regions are distributed in a particular way. In some plants including wheat (Triticum aestivum), maize (Zea mays) and barley (Hordeum vulgare) the crossover frequency tends to increase with the relative physical distance from the centromere. By contrast, in Allium stulosum (Welsh onion), crossovers seem to cluster close to centromeres [3]. For other plants for example, Arabidopsis, rice (Oryza sativa)

55

Review

TRENDS in Genetics

Vol.23 No.2

93

Figure 1. Prophase I of meiosis. (a) The cytological progression of prophase I of meiosis. After replication, chromosomes condense into long threads. This stage is known as leptotene. Then, at zygotene, the pairs of homologous chromosomes start to synapse, through formation of the synaptonemal complex. Next, at pachytene, synapsis is completed, and the synaptonemal complex links the homologs along their whole length. Then, at diakinesis (the final stage of prophase I), the synaptonemal complex has disappeared, and the further condensation of chromosomes reveals the presence of chiasmata. (b) A model of meiotic recombination pathways in Saccharomyces cerevisiae (yeast) and Arabidopsis thaliana. DNA DSBs are formed at leptotene and are processed to form single strands, which then invade a chromatid of the homologous chromosome as a repair template. Repair results in three different products: interfering crossovers, noninterfering crossovers and noncrossovers. Proteins involved in this process are indicated, together with the stage they are thought to operate at; data taken from Ref. [78]. Question marks indicate Arabidopsis proteins that are thought to be involved in this process on the basis of their similarity with S. cerevisiae proteins but that have not been functionally characterized. Alternative names for S. cerevisiae and Arabidopsis proteins are indicated here in parentheses: Rec107 (Mer2), Ski8 (Rec103), Mre2 (Nam8), Sae2 (Com1), Zip4 (Spo22), AtMSH5 (At3G20475), AtMUS81A (At4G30870) and AtMUS81B (At5G39770).

and tomato (Solanum lycopersicum) crossover distribution varies between and along chromosome arms, with no apparent rule. Interestingly, within regions (cold or hot), crossover rates vary enormously from one kilobase
www.sciencedirect.com

to another in Arabidopsis [20] (Figure 2). These observations led to the hypothesis that there are several levels of control, each operating at a different scale: chromosomal, regional (megabase) and local (kilobase).

56

94

Review

TRENDS in Genetics

Vol.23 No.2

Figure 2. Variation in crossover rate along chromosome 4 of Arabidopsis thaliana. (a) The graph shows the variation in crossover rate along chromosome 4 of Arabidopsis. A schematic representation of chromosome 4 is aligned with the graph, showing the centromere (red), the heterochromatic knob (blue) and the nucleolar organizer region (yellow). Part (a) reproduced, with permission, from Ref. [79] (2006) Cold Spring Harbor Laboratory Press. (b) The graph shows the variation in crossover rate in the first 800 kb of the short arm of chromosome 4. Crossover breakpoints are concentrated in ten regions of <5 kb on this short arm, resulting in a recombination rate at least fivefold greater than the whole-chromosome average [4.6 centimorgan (cM)/Mb; dashed red line in (a)]. Part (b) reproduced, with permission, from Ref. [20] (2006) the Biochemical Society.

In contrast to crossovers, the density of early recombination events, marked by ENs, is less variable along chromosomes [15,22]. ENs, or foci of either Rad51 or Dmc1, are often much more numerous than crossovers, strongly suggesting that DSBs are mostly repaired by noncrossovers. For example, there are 40-fold more ENs than LNs in Allium species [14], and there are tenfold more Rad51 or Dmc1 foci than crossovers in the mouse [23]. In Arabidopsis, the ratio of AtRAD51 foci to crossovers is $25:1 [24], and models developed to infer genetic recombination rates from population genetic data t better with a 1:1 ratio of gene conversion to crossovers [25]. The 1:1 estimate needs to be taken with caution, because it could be biased by several parameters (e.g. mutation and recombination rates, and sampling). In addition, this analysis detects only gene conversion and not silent noncrossovers. The AtRAD51 foci probably provide a better estimate of the noncrossover to crossover ratio, even if they might mark other events (e.g. SCE). The EN to LN ratio varies from interval to interval (from 4:1 to 20:1 in maize) [14,26], indicating that all early recombination events have different probabilities of becoming a crossover. This point is also illustrated by
www.sciencedirect.com

the noncrossover to crossover ratio varying substantially from one locus to another in organisms as different as yeast and humans [2729]. Therefore, the variation of crossover frequencies along a chromosome depends greatly on the choice of DSB-repair mechanism (crossover versus noncrossover), not only on the DSB distribution per se. Interference shapes the distribution of crossovers In addition to being unevenly distributed along chromosomes, crossovers are not independent of each other, because of the phenomenon of interference. Interference was dened by Muller in 1916 [30]: The occurrence of one crossing-over interferes with the coincident occurrence of another crossing-over in the same pair of chromosomes, and I have accordingly termed this phenomenon interference. An obvious consequence of interference is that crossovers are more distant from each other than would be expected if they were independent. Interference is commonly estimated genetically by the coefcient of coincidence [30] or cytologically by the distribution of LNs or chiasmata [31] (Box 2). The existence of interference has been conrmed in most species that have been tested; exceptions include Schizosaccharomyces pombe

57

Review

TRENDS in Genetics

Vol.23 No.2

95

Box 2. The measurement of interference: C and n


The first objective measurement of crossover interference, the coefficient of coincidence (C), was described by Muller [30]. It is the ratio of the observed frequency of crossovers being simultaneously present in two different intervals to the product of the frequency of crossovers in each interval. Its value ranges from 1, when the occurrence of a crossover has no effect on the occurrence of another crossover (i.e. no interference), to 0, when the presence of a crossover prohibits the occurrence of another crossover (i.e. total interference). Between these values, C indicates the strength of interference. There are several reports of negative interference (C >1). We suggest that this type of interference falls into two categories. At short distances (a few kilobases), a noncrossover can be misinterpreted as two crossovers. When observed between two distant intervals, negative interference could be a side-effect of positive interference, resulting from an increased probability of having a second crossover at a distance corresponding to the end of the effect of positive interference [76,77]. Plotting C between each pair of intervals as a function of genetic distance yields a coincidence function that gives an overall view of interference [77] (as modeled in Figure Ib). Among several mathematical models that have been elaborated to describe this coincidence function, gamma models turned out to provide the best fit between experimental and predicted C values [77]. Indeed, gamma models can be used to provide a distribution of the distances between crossovers (i.e. the inter-crossover distances; Figure Ia) from which a coincidence function can be deduced (Figure Ib). The probability density function of the gamma distribution in Figure Ia is shown in Equation I: f x ; n; l ln x n1 e lx G n [I] Therefore, n is a measure of the strength of interference. It can be estimated either by finding the n value that provides the best fit for the observed distribution of intercrossover distances (measured genetically or cytologically [58]) or from observed C values [64]. The gamma model can be used as a tool to measure interference without assuming any particular biological mechanism [58].

where x is the input variable, n is a shape parameter, l is a rate parameter, and G(n) is the gamma function; the mean is n/l, and the variance is n/l2. The shape of a gamma distribution curve is described by the parameter n [58], which, biologically, reflects the number of noncrossovers between two crossovers plus one (m + 1). When n is 1, the distribution of the intercrossover distances is exponential (Figure Ia), and the coincidence function is a horizontal straight line that has a value of 1 (Figure Ib). That is, for n = 1, intercrossover distances are random, so there is no interference. As n increases, the distribution of the intercrossover distances tends to be gaussian (Figure Ia), and the C values in the deduced coincidence functions decrease at short distances (Figure Ib). In other words, for high values of n, the intercrossover distances are constrained and become less variable, and interference is strong.

Figure I. Using gamma distributions to determine the strength of interference. (a) Gamma distributions with varying shape (n) and rate (l) parameters and a constant mean. (b) Coincidence functions deduced from gamma distributions with a varying shape (n) parameter.

(ssion yeast) and Aspergillus nidulans (also known as Emericella nidulans) [15]. For example, in tomato and maize, the observed distribution of the distance between two LNs is different from that expected if their positions were assumed to be independently and homogeneously distributed [31,32]. In addition, detailed analyses of crossovers in yeast, humans and Arabidopsis strongly indicate that although most crossovers are subject to interference, some are not [3337]. The existence of two classes of crossover (one subject to interference and one not) has been conrmed by mutant analyses in yeast [1,2] and Arabidopsis [3841]. Formation of interfering crossovers In yeast, interfering-crossover formation is catalyzed by the recruitment of a set of meiosis-specic proteins referred to as ZMM proteins [for Zip1, Zip2, Zip3 (also known as Cst9), Msh4, Msh5 and Mer3 (also known as Hfm1) proteins] [42] (Figure 1) to a subset of DSB sites. Noninterfering crossovers depend on the proteins Mus81
www.sciencedirect.com

and Mms4 [43]. The actions of ZMM proteins indicate that interfering-crossover sites are selected from potential precursors at an early stage in the meiotic DSB-repair process [1,44]. In Arabidopsis, the depletion of AtMSH4 and AtMER3 (also known as MER3), two of the ZMM proteins, greatly reduces the number of crossovers that form, with atmsh4 mutants having 85% fewer chiasmata than wildtype Arabidopsis [39] and atmer3 mutants having 75% fewer [38,40]. The interference between the crossovers that are present is substantially reduced in atmer3 mutants [40]. Therefore, the interfering pathway is responsible for the formation of most crossovers in Arabidopsis and involves AtMSH4 and AtMER3. Copenhaver et al. also estimated a proportion of $15% of noninterfering crossovers by tting model data to genetic data [45]. Zip1, another ZMM protein in yeast, has a structural role in the synaptonemal complex [42,46]. It had been proposed that the polymerization of the synaptonemal complex could be the interference signal, but there is now strong evidence against the idea of such a mechanism

58

96

Review

TRENDS in Genetics

Vol.23 No.2

[1,47]. Unlike the situation in yeast, in which the absence of Zip1 leads to a marked reduction in the number of crossovers, crossover formation is only slightly reduced (by 20%) when the two Arabidopsis orthologs of yeast ZIP1 (ATZYP1A and ATZYP1B; also known as ZYP1A and ZYP1B) are inactivated using RNA interference [48]. Because no direct assay was carried out to study interference in this context, additional analysis is needed to conrm or to disprove the role of the AtZYP1 proteins in interference in Arabidopsis. In addition, multivalents and bivalents between nonhomologous chromosomes are observed when both ATZYP1 genes are inactivated, showing that the AtZYP1 proteins are required to restrain the formation of crossovers between homologous chromosomes [48]. These results indicate that, in Arabidopsis, AtZYP1A and AtZYP1B are not ZMM proteins as dened in yeast. In yeast and Arabidopsis, other proteins also promote crossover formation (Figure 1). Meiotic crossover defects have been described in mlh1 [49] and mlh3 [50] deletion mutants in yeast and in atmlh3 [51] mutants in Arabidopsis but seem to be less severe than those in yeast msh4 or Arabidopsis atmsh4 mutants. The role of these factors in interference is either controversial (for yeast mlh1 and mlh3) [52] or unconrmed (for Arabidopsis atmlh3), and their relationship to ZMM proteins is still unclear. In Arabidopsis, PARTING DANCERS (ATPTD; also known as AT1G12790.1), a novel gene that has no homolog in other kingdoms, might function in the interfering pathway: atptd mutants have 75% fewer chiasmata than wild-type Arabidopsis [41]. The list of genes implicated in crossover formation is growing, and it is unlikely that all the factors involved have been identied. Three of the proteins involved in the formation of interfering crossovers have been immunolocalized in several species. In the mouse, MLH1 and MLH3 localize at crossover sites on meiotic chromosomes [53] and are commonly used to count crossovers, although it is unknown whether they mark all crossover sites. In tomato, there are 30% fewer MLH1 foci than LNs, and comparisons of their distributions indicate that interference among MLH1 foci is much stronger than among LNs. Because LNs correspond closely to chiasmata [31], this nding indicates that LNs mark all sites of crossovers (interfering and noninterfering), whereas MLH1 foci mark only a subset of strongly interfering crossovers. In addition, simulation shows that the distribution obtained when adding 30% more crossovers, in random positions, to the observed distribution of MLH1 foci matches the observed distribution of LNs [54] (C. Heyting, unpublished). In Arabidopsis, AtMLH1 and AtMLH3 foci colocalize at pachytene [51]. These foci are slightly less numerous than the chiasmata count at metaphase I, which itself is probably a small underestimation of the number of crossovers. This suggests that these two proteins mark the sites of interfering crossovers and not the sites of noninterfering crossovers as in tomato. The situation is different for Msh4 and its orthologs in other species. In yeast, the number of Msh4 foci matches the number of crossovers [55]. By contrast, in the mouse and in Arabidopsis, numerous AtMSH4 foci appear at mid-leptotene, clearly marking more sites than there are crossovers [39,56]; the function of Msh4 and its
www.sciencedirect.com

orthologs at these sites still needs to be elucidated. Then, the number of these sites gradually decreases at zygotene and pachytene. In Arabidopsis, the colocalization of AtMLH3 and AtMSH4 is poor at pachytene [51], indicating that these proteins do not function at the same time in the formation of interfering crossovers. Noninterfering crossovers coexist with interfering crossovers Two factors in the noninterfering-crossover pathway have been identied in yeast: Mus81 and Mms4 [43] (Figure 1). These proteins are responsible for $30% of all crossovers. Arabidopsis contains two putative homologs of MUS81 [57], which might function in the noninterfering-crossover pathway. The ratio of interfering crossovers to noninterfering crossovers differs among the species in which it has been studied. In yeast and tomato, $30% of crossovers seem to escape the interference mechanism; and in Arabidopsis, 15%. At the two extremes are Caenorhabditis elegans, which has only interfering crossovers, and S. pombe (which lacks Msh4 and Msh5), which has only noninterfering crossovers [2]. Interestingly, in plants as diverse as potato, poplar, rice and Arabidopsis, homologs of the two sets of proteins (those involved in interfering crossovers and those involved in noninterfering crossovers) can be found, suggesting that the coexistence of the two pathways is a common feature in the plant kingdom. It is unclear why certain species have conserved both pathways and why the balance between the two pathways differs markedly. To add to this complexity, some crossovers still occur when both interfering-crossover (ZMM proteins) and noninterfering-crossover (Mus81 or Mms4) pathways are simultaneously disrupted in yeast [43,50], suggesting the existence of a third crossover-formation mechanism. Are there several layers of interference? That interference occurs between LNs, crossovers or Mlh1 foci at pachytene is no longer in doubt. Interestingly, a low level of interference has been detected between MSH4 foci (half that at MLH1 foci) at late zygotene in the mouse [58]. The distribution of ENs in tomato seems to show interference [58], in contrast to previous descriptions of a large range of plant species, for which the distribution of ENs was reported to be free of interference (except possibly over short distances) [22]. These new ndings indicate that there are also two levels of interference in tomato. It is unknown whether these two putative levels of interference are mediated by two independent, still undened, mechanisms and whether they are concomitant or are established successively as meiosis progresses. If ENs mark all DSB sites, then the rst level of interference must be at the stage of DSB formation. At present, although it is clear that, in yeast, the density of DSBs varies enormously along chromosomes [59,60], there are no data about putative interference between DSBs. Models of interference As it became clear that crossovers and noncrossovers both arise from meiotic DSBs and have a common early-recombination DNA structure, it was postulated that

59

Review

TRENDS in Genetics

Vol.23 No.2

97

Figure 3. The process of crossover formation. The start point of crossover formation is a distribution of precursors (i.e. DSBs) among and along chromosomes (a). The density of these precursors varies along chromosomes, but how their distribution is controlled is largely unknown. These DSBs can be repaired as crossovers or noncrossovers. The tendency of individual precursors to become a crossover probably varies [28,29]. The mechanism that controls the decision of whether a precursor is repaired as a crossover or noncrossover is unknown. When one of the precursors, assumed to be stochastically chosen, is designated to become a crossover (b), this site inhibits the formation of a second crossover in its vicinity, implying that precursors subject to this effect become noncrossovers (c). The outcome is shown in (d). The way in which potential crossovers crosstalk to establish such interference at the chromosome level is not understood. The precursors located outside the effect of interference (i.e. the interference signal) might not become crossovers (i), or they can be designated as a second crossover and, in turn, inhibit the formation of other crossovers nearby (ii,iii). This controlled selection of a few sites to become crossovers from a large population of precursors (assuring that each bivalent has at least one crossover) explains the two main properties of crossover distribution: the obligatory crossover and interference. (For a discussion of the relationship between obligatory chiasmata and interference, see Ref. [80].) A proportion of precursors can be designated to become noninterfering crossovers, independent of the presence of another crossover nearby (iii). The way in which some precursors enter this pathway, and why they are not subject to interference, is still puzzling.

interference is imposed when they diverge [61] (Figure 3). This led to the proposal of a series of models to explain the control of the differentiation between crossovers and noncrossovers and the coordination at the chromosome level of this control. In the counting model, the number of noncrossover events between crossovers (m) is xed. Therefore, m directly reects the strength of interference (Box 2). The t of the counting model to genetic data sets is satisfactory in most cases [62], so this became one of the main models in the eld. However, the basis for xing the value of m is challenged by numerous data indicating that m varies between sexes and individuals in the same species, between chromosomes in the same nucleus and even along chromosomes [63,64]. To account for the variations in strength of interference, two modied versions of the counting model were proposed: one in which m is not constant [65], and a second in which the existence of noninterfering crossovers is taken into account [33,35 37]. In this second model, the apparent variation in interference strength is due to the variable proportion of
www.sciencedirect.com

noninterfering crossovers to interfering crossovers ( p) from chromosome to chromosome and along chromosomes, rather than to variation in the mechanism of interference itself (m). Indeed, this assumption improves the t of the model to the genetic data from yeast, humans and Arabidopsis [33,3537]. However, the best tting values for m and p in Arabidopsis [33,37] predict that the crossovers that occur in ZMM protein mutants will not be distributed equally among chromosomes. This prediction is not substantiated in atmsh4 [39] and atmer3 mutants (J. Vignard and R. Mercier, unpublished), indicating that a variation in the ratio of noninterfering crossovers to interfering crossovers is insufcient to account for the variation in interference between chromosomes. Another set of data obtained recently in yeast also presents serious challenges to the counting model. Indeed, the decrease in the number of meiotic DSBs was shown not to reduce the number of crossovers proportionally (a phenomenon known as crossover homeostasis) and not to modify the strength of interference [66]. This contradicts the prediction of the counting model, which nds that to keep the number of

60

98

Review

TRENDS in Genetics

Vol.23 No.2

noncrossovers constant between two crossovers (m), the number of crossovers needs to decrease. Another model postulates that early recombination intermediates sometimes become crossovers and then nucleate an inhibitory polymerization; this interference signal then propagates in its own vicinity and forces other early events to resolve as noncrossovers [67]. The more obvious candidate for the propagation of this signal was the polymerization of the synaptonemal complex, but there is now strong evidence against the idea of such a mechanism [1,47]. No other potential mediator of this signal has been identied. More recently, in another model, known as the mechanical stress model, the designation of crossovers among DSBs results from the imposition of a physical stress on the chromatin ber [68]. The occurrence of a crossover results in local relief of the stress, which spreads out in both directions from the affected site with decreasing strength. Therefore, the probability of a second crossover occurring adjacent to the rst is zero and progressively increases with distance, up to the initial probability. A simulation according to this model successfully tted genetic data from two species. However, there has been no in vivo demonstration supporting this model. Conclusions As more genetic and cytological data become available, in particular data from mutants of various species, the emerging picture of crossover formation is becoming more detailed and is turning out to be more complex than was expected a few years ago. The nal distribution of crossovers results from the integration of several levels of control (Figure 3), which can be summarized as follows: (i) the density of DSBs varies along chromosomes in yeast, and probably in other species, and DSBs might interfere; (ii) the propensity of a DSB to become a crossover or a noncrossover (or an SCE) probably varies; (iii) interference might operate at several levels, and its strength varies along a chromosome and among chromosomes; (iv) some crossovers, the proportion of which differs among species, are insensitive to interference; and (v) there might be more than two pathways controlling crossover formation. At present, each of these different levels of control of crossover formation is poorly understood, and there is still no unied theory of crossover differentiation that integrates genetic and cytological data and proposes a satisfactory underlying molecular mechanism. However, there is much research activity in the eld, and communication between groups working on a large range of model species in different kingdoms has recently become intense. Plants have their say in this dialog thanks to the combination of available genetic data and superb cytological tools for studying several species, in addition to the molecular mechanism of crossover formation having been deciphered, particularly in Arabidopsis. This multi-organism approach should soon result in new insights into the mechanisms involved in each level of control of the nal crossover distribution.
Acknowledgements
We thank Franc oise Budar, Mathilde Grelon, Eric Jenczewski and Olivier Loudet for critical reading of the manuscript. Many thanks also to Christa Heyting for sharing data before publication.
www.sciencedirect.com

References
1 Bishop, D.K. and Zickler, D. (2004) Early decision; meiotic crossover interference prior to stable strand exchange and synapsis. Cell 117, 9 15 2 Hollingsworth, N.M. and Brill, S.J. (2004) The Mus81 solution to resolution: generating meiotic crossovers without Holliday junctions. Genes Dev. 18, 117125 3 Jones, G.H. (1984) The control of chiasma distribution. Symp. Soc. Exp. Biol. 38, 293320 4 Lynn, A. et al. (2004) Variation in human meiotic recombination. Annu. Rev. Genomics Hum. Genet. 5, 317349 5 Bateson, W. (1909) Mendels Principles of Heredity, Cambridge University Press 6 Keeney, S. (2001) Mechanism and control of meiotic recombination initiation. Curr. Top. Dev. Biol. 52, 153 7 Schwacha, A. and Kleckner, N. (1997) Interhomolog bias during meiotic recombination: meiotic functions promote a highly differentiated interhomolog-only pathway. Cell 90, 11231135 8 Grelon, M. et al. (2001) AtSPO11-1 is necessary for efcient meiotic recombination in plants. EMBO J. 20, 589600 9 Stacey, N.J. et al. (2006) Arabidopsis SPO11-2 functions with SPO11-1 in meiotic recombination. Plant J. 48, 206216 10 Bleuyard, J.Y. et al. (2004) Meiotic defects in the Arabidopsis rad50 mutant point to conservation of the MRX complex function in early stages of meiotic recombination. Chromosoma 113, 197203 11 Puizina, J. et al. (2004) Mre11 deciency in Arabidopsis is associated with chromosomal instability in somatic cells and Spo11dependent genome fragmentation during meiosis. Plant Cell 16, 1968 1978 12 Jolivet, S. et al. (2006) Non conservation of the meiotic function of the Ski8/Rec103 homolog in Arabidopsis. Genes Cells 11, 615622 13 Stack, S.M. et al. (1989) Chiasmata and recombination nodules in Lilium longiorum. Genome 32, 486498 14 Anderson, L.K. and Stack, S.M. (2005) Recombination nodules in plants. Cytogenet. Genome Res. 109, 198204 15 Zickler, D. and Kleckner, N. (1999) Meiotic chromosomes: integrating structure and function. Annu. Rev. Genet. 33, 603754 16 Anderson, L.K. et al. (1997) RecA-like proteins are components of early meiotic nodules in lily. Proc. Natl. Acad. Sci. U. S. A. 94, 68686873 17 Li, W. et al. (2004) The Arabidopsis AtRAD51 gene is dispensable for vegetative development but required for meiosis. Proc. Natl. Acad. Sci. U. S. A. 101, 1059610601 18 Couteau, F. et al. (1999) Random chromosome segregation without meiotic arrest in both male and female meiocytes of a dmc1 mutant of Arabidopsis. Plant Cell 11, 16231634 19 Siaud, N. et al. (2004) Brca2 is involved in meiosis in Arabidopsis thaliana as suggested by its interaction with Dmc1. EMBO J. 23, 1392 1401 20 Mezard, C. (2006) Meiotic recombination hot spots in plants. Biochem. Soc. Trans. 34, 531534 21 Choo, K.H. (1998) Why is the centromere so cold? Genome Res. 8, 8182 22 Anderson, L.K. et al. (2001) The distribution of early recombination nodules on zygotene bivalents from plants. Genetics 159, 1259 1269 23 Moens, P.B. et al. (2002) The time course and chromosomal localization of recombination-related proteins at meiosis in the mouse are compatible with models that can resolve the early DNADNA interactions without reciprocal recombination. J. Cell Sci. 115, 16111622 24 Chelysheva, L. et al. (2005) AtREC8 and AtSCC3 are essential to the monopolar orientation of the kinetochores during meiosis. J. Cell Sci. 118, 46214632 25 Plagnol, V. et al. (2005) Relative inuences of crossing-over and gene conversion on the pattern of linkage disequilibrium in Arabidopsis thaliana. Genetics 172, 24412448 26 Stack, S.M. and Anderson, L.K. (2002) Crossing over as assessed by late recombination nodules is related to the pattern of synapsis and the distribution of early recombination nodules in maize. Chromosome Res. 10, 329345 27 Fogel, S. et al. (1981) Mechanisms of meiotic gene conversion, or wanderings on a foreign strand. In The Molecular Biology of the Yeast Saccharomyces: Life Cycle and Inheritance (Strathern, J. et al., eds), pp. 289339, Cold Spring Harbor Laboratory Press

61

Review

TRENDS in Genetics

Vol.23 No.2

99

28 Jeffreys, A.J. and Neumann, R. (2005) Factors inuencing recombination frequency and distribution in a human meiotic crossover hotspot. Hum. Mol. Genet. 37, 601606 29 Kauppi, L. et al. (2004) Where the crossovers are: recombination distributions in mammals. Nat. Rev. Genet. 5, 413424 30 Muller, H.J. (1916) The mechanisms of crossing-over. Am. Nat. 50, 193221 31 Sherman, J.D. and Stack, S.M. (1995) Two-dimensional spreads of synaptonemal complexes from solanaceous plants. VI. High-resolution recombination nodule map for tomato (Lycopersicon esculentum). Genetics 141, 683708 32 Anderson, L.K. et al. (2003) High-resolution crossover maps for each bivalent of Zea mays using recombination nodules. Genetics 165, 849865 33 Lam, S.Y. et al. (2005) Crossover interference on nucleolus organizing region-bearing chromosomes in Arabidopsis. Genetics 170, 807812 34 Malkova, A. et al. (2004) Gene conversion and crossing over along the 405-kb left arm of Saccharomyces cerevisiae chromosome VII. Genetics 168, 4963 35 Stahl, F.W. et al. (2004) Does crossover interference count in Saccharomyces cerevisiae? Genetics 168, 3548 36 Housworth, E.A. and Stahl, F.W. (2003) Crossover interference in humans. Am. J. Hum. Genet. 73, 188197 37 Copenhaver, G.P. et al. (2002) Crossover interference in Arabidopsis. Genetics 160, 16311639 38 Chen, C. et al. (2005) The Arabidopsis ROCK-N-ROLLERS gene encodes a homolog of the yeast ATP-dependent DNA helicase MER3 and is required for normal meiotic crossover formation. Plant J. 43, 321334 39 Higgins, J.D. et al. (2004) The Arabidopsis MutS homolog AtMSH4 functions at an early step in recombination: evidence for two classes of recombination in Arabidopsis. Genes Dev. 18, 25572570 40 Mercier, R. et al. (2005) Two meiotic crossover classes cohabit in Arabidopsis: one is dependent on MER3, whereas the other one is not. Curr. Biol. 15, 692701 41 Wijeratne, A.J. et al. (2006) The Arabidopsis thaliana PARTING DANCERS gene encoding a novel protein is required for normal meiotic homologous recombination. Mol. Biol. Cell 17, 13311343 42 Borner, G.V. et al. (2004) Crossover/noncrossover differentiation, synaptonemal complex formation, and regulatory surveillance at the leptotene/zygotene transition of meiosis. Cell 117, 2945 43 De Los Santos, T. et al. (2003) The Mus81/Mms4 endonuclease acts independently of double-Holliday junction resolution to promote a distinct subset of crossovers during meiosis in budding yeast. Genetics 164, 8194 44 Fung, J.C. et al. (2004) Imposition of crossover interference through the nonrandom distribution of synapsis initiation complexes. Cell 116, 795802 45 Copenhaver, G.P. et al. (1998) Assaying genome-wide recombination and centromere functions with Arabidopsis tetrads. Proc. Natl. Acad. Sci. U. S. A. 95, 247252 46 Sym, M. et al. (1993) ZIP1 is a synaptonemal complex protein required for meiotic chromosome synapsis. Cell 72, 365378 47 de Boer, E. and Heyting, C. (2006) The diverse roles of transverse laments of synaptonemal complexes in meiosis. Chromosoma 115, 220234 48 Higgins, J.D. et al. (2005) The Arabidopsis synaptonemal complex protein ZYP1 is required for chromosome synapsis and normal delity of crossing over. Genes Dev. 19, 24882500 49 Wang, T.F. et al. (1999) Functional specicity of MutL homologs in yeast: evidence for three Mlh1-based heterocomplexes with distinct roles during meiosis in recombination and mismatch correction. Proc. Natl. Acad. Sci. U. S. A. 96, 1391413919 50 Argueso, J.L. et al. (2004) Competing crossover pathways act during meiosis in Saccharomyces cerevisiae. Genetics 168, 18051816 51 Jackson, N. et al. (2006) Reduced meiotic crossovers and delayed prophase I progression in AtMLH3-decient Arabidopsis. EMBO J. 25, 13151323 52 Hoffmann, E.R. and Borts, R.H. (2004) Meiotic recombination intermediates and mismatch repair proteins. Cytogenet. Genome Res. 107, 232248

53 Marcon, E. and Moens, P. (2003) MLH1p and MLH3p localize to precociously induced chiasmata of okadaic-acid-treated mouse spermatocytes. Genetics 165, 22832287 54 Lhuissier, F.G.P. et al. (2005) An immunocytological recombination map of tomato. In The European Meiosis Meeting, p. 104, European Molecular Biology Organization 55 Novak, J.E. et al. (2001) The budding yeast Msh4 protein functions in chromosome synapsis and the regulation of crossover distribution. Genetics 158, 10131025 56 Kneitz, B. et al. (2000) MutS homolog 4 localization to meiotic chromosomes is required for chromosome pairing during meiosis in male and female mice. Genes Dev. 14, 10851097 57 Hartung, F. et al. (2006) The role of AtMUS81 in DNA repair and its genetic interaction with the helicase AtRecQ4A. Nucleic Acids Res. 34, 44384448 58 de Boer, E. et al. (2006) Two levels of interference in mouse meiotic recombination. Proc. Natl. Acad. Sci. U. S. A. 103, 9607 9612 59 Gerton, J.L. et al. (2000) Global mapping of meiotic recombination hotspots and coldspots in the yeast Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. U. S. A. 97, 1138311390 60 Baudat, F. and Nicolas, A. (1997) Clustering of meiotic double-strand breaks on yeast chromosome III. Proc. Natl. Acad. Sci. U. S. A. 94, 52135218 61 Mortimer, R.K. and Fogel, S. (1974) Genetical interference and gene conversion. In Mechanism in Recombination (Grell, R.F., ed.), pp. 263 275, Plenum Press 62 Foss, E. et al. (1993) Chiasma interference as a function of genetic distance. Genetics 133, 681691 63 Esch, E. (2005) Estimation of gametic frequencies from F2 populations using the EM algorithm and its application in the analysis of crossover interference in rice. Theor. Appl. Genet. 111, 100109 64 Lin, S. et al. (2001) Genetic crossover interference in the human genome. Ann. Hum. Genet. 65, 7993 65 Lange, K. et al. (1997) The Poisson-skip model of crossing-over. Ann. Appl. Probab. 7, 299313 66 Martini, E. et al. (2006) Crossover homeostasis in yeast meiosis. Cell 126, 285295 67 King, J.S. and Mortimer, R.K. (1990) A polymerization model of chiasma interference and corresponding computer simulation. Genetics 126, 11271138 68 Kleckner, N. et al. (2004) A mechanical basis for chromosome function. Proc. Natl. Acad. Sci. U. S. A. 101, 1259212597 69 Mendel, G. (1866) Versuche u ber panzenhybriden. Verh. Naturf. Ver. Bru nn. 4, 3-44 70 Whitehouse, H.L.K. (1973) Towards an Understanding of the Mechanisms of Heredity, (3rd edn), Edward Arnold 71 Bateson, W. et al. (1905) Experimental studies in the physiology of heredity. Rep. Evol. Comm. R. Soc. 2, 1154 72 Creighton, H.B. and McClintock, B. (1931) A correlation of cytological and genetical crossing-over in Zea mays. Proc. Natl. Acad. Sci. U. S. A. 17, 492497 73 Haldane, J.B.S. (1931) The cytological basis of genetical interference. Cytologia (Tokyo) 3, 5465 74 McClintock, B. (1956) Controlling elements and the gene. Cold Spring Harbor Symp. Quant. Biol. 21, 197216 75 Vaucheret, H. (2006) Post-transcriptional small RNA pathways in plants: mechanisms and regulations. Genes Dev. 20, 759771 76 Colombo, P.C. and Jones, G.H. (1997) Chiasma interference is blind to centromeres. Heredity 79, 214227 77 McPeek, M.S. and Speed, T.P. (1995) Modeling interference in genetic recombination. Genetics 139, 10311044 78 Hamant, O. et al. (2006) Genetics of meiotic prophase I in plants. Annu. Rev. Plant Biol. 57, 267302 79 Drouaud, J. et al. (2006) Variation in crossing-over rates across chromosome 4 of Arabidopsis thaliana reveals the presence of meiotic recombination hot spots. Genome Res. 16, 106114 80 Jones, G.H. and Franklin, F.C. (2006) Meiotic crossing-over: obligation and interference. Cell 126, 246248

www.sciencedirect.com

62

2.2 Distributions de crossovers spcifiques du sexe et variations du niveau de linterfrence le long du chromosome 4 dArabidopsis thaliana
2.2.1 Synthse de la publication
Afin dtudier la distribution des CO le long du chromosome 4 dA. thaliana, sparment dans les mioses des deux sexes, deux populations denviron 1350 plantes ont t construites, issues du rtrocroisement dun hybride F1 entre les accessions Columbia-0 et Landsberg erecta-1, par laccession Columbia-0 utilise comme parent mle ou femelle. Ces populations ont ensuite t gnotypes en utilisant un jeu de 43 marqueurs SNP couvrant lensemble du chromosome 4 (18,6 Mb ; taille moyenne des intervalles 423 kb). Le nombre moyen de crossings-over par miose atteint 1,76 dans la miose mle, contre 1,05 dans la miose femelle. Cela signifie que la proportion de crossings-over multiples dans la miose femelle est faible : seul le crossing-over obligatoire est gnralement prsent. Cette htrochiasmie est corrle la diffrence de taille totale du complexe synaptonmal entre les sexes : le rapport mle/femelle des longueurs de CS atteint 1,69 tandis que le rapport des nombres de crossings-over est gal 1,68. Les taux de CO diffrent beaucoup entre les mioses mle et femelle aux deux extrmits du chromosome, mais voluent paralllement sans cart significatif sur une large portion centrale. La rgion pricentromrique, et le knob htrochromatinien adjacent sur le bras court, sont dpourvus de CO. Linterfrence des CO a ensuite t tudie dans les mioses des deux sexes. Aussi bien la distribution du nombre de CO par chromatide, que la distribution des distances qui sparent les CO successifs, dmontrent lexistence de linterfrence. La concidence quatre points des CO prsents dans deux intervalles disjoints augmente avec la distance gntique qui les spare. Dans la miose mle, elle atteint voire dpasse 1 pour les intervalles les plus loigns, ce qui signifie que linterfrence sannule, puis devient lgrement ngative une distance donne : il existe alors une frquence plus forte quattendue de CO doubles. La taille gntique du chromosome 4 en miose femelle est trop faible pour pouvoir observer cet effet. La concidence 3 points entre intervalles adjacents, et de taille gntique fixe, a galement t examine. Dans la miose mle, elle varie de faon trs significative le long du chromosome, mais apparemment pas dans la miose femelle (cela peut tre du la moindre 63

Population mle % des % des Crossovers/chromatide ou chromatides bivalents chiasmas/bivalent 0 32 -1,3 1 49,4 37,1 2 17,3 54,1 3 1,3 10,1 total 100 100

Population femelle % des % des chromatides bivalents 50 3,9 47,9 88,7 2 6,1 0,2 1,2 100 100

Table 5. Distributions du nombre de crossovers par chromatide et du nombre de chiasmas par bivalent dans les mioses mle et femelle chez Arabidopsis thaliana .

Chiasmas/bivalent 1 0 1 2 3 total Seuil Khi-carr

Population mle Bivalents Bivalents observs attendus 127 169 192 125 36 47 0 14 355 355 2,78.10-5

Population femelle Bivalents Bivalents observs attendus 302 299 20 26 4 1 0 0 326 326 0,31

Table 6. Distributions observe et attendue du nombre de chiasmas nonobligatoires par bivalent dans les mioses mle et femelle chez Arabidopsis thaliana .

quantit de donnes). La variation de la concidence des CO observe dans la miose mle est corrle la taille physique des paires dintervalles considrs. Cela signifie que pour une frquence totale de CO donne, la frquence de CO doubles augmente avec la distance physique qui les spare.

2.2.2 Rsultats additionnels


La distribution thorique du nombre de crossings-over par bivalent a t calcule en utilisant la formule de Mather (Mather 1938), sous lhypothse dune absence dinterfrence chromatidienne (table 5). On constate que les bivalents pourvus de deux crossings-over sont majoritaires dans la miose mle, tandis que les bivalents de la miose femelle nen prsentent gnralement quun seul. La proportion de bivalents dpourvus de crossings-over est ngative dans la population mle. Cela peut sexpliquer par la prsence dinterfrence chromatidienne, mais plus vraisemblablement par une variation stochastique des effectifs due lchantillonnage, tant donne la faiblesse de lcart. En revanche, la proportion prdite de bivalents achiasmatiques dans la miose femelle est plus importante, et semble ne pas pouvoir tre explique par la seule variance dchantillon. Il semble donc possible que la loi du crossing-over obligatoire ne soit pas toujours respecte dans la miose femelle. Cependant, si ce taux de 4% sappliquait galement aux cinq bivalents, cela devrait aboutir la formation de spores aneuplodes dans 10 % des cas, donc 10 % de ltalit embryonnaire, ce qui nest jamais observ. Il est donc vraisemblable que la frquence de bivalents 4 achiasmatiques, bien que significativement non nulle, soit en fait infrieure 4%. Si le crossing-over obligatoire est t de chacune des classes, la distribution des crossings-over restants est conforme une distribution de Poisson dans le cas de la miose femelle, mais pas dans celui de la miose mle (table 6). Cela peut signifier que les bivalents de la miose femelle forment dans 96 % des cas au moins un crossing-over, et des crossings-over dune faon alatoire dans 7,4 % des cas. La distribution du nombre de crossings-over dans la miose mle ne se prte pas ce genre de simplification, car linterfrence ntant pas totale, il nest pas possible de distinguer les crossings-over des deux voies.

2.2.3 Publication

64

Sex-Specific Crossover Distributions and Variations in Interference Level along Arabidopsis thaliana Chromosome 4
l Mercier1[, Liudmila Chelysheva1, Aure lie Be rard2, Matthieu Falque3, Olivier Martin3,4, Jan Drouaud1[, Raphae 1 2 1* zard Vanessa Zanni , Dominique Brunel , Christine Me
ne tique et dAme lioration des Plantes, Institut Jean Pierre Bourgin, INRA, Versailles, France, 2 UR Etude du Polymorphisme des Ge nomes Ve ge taux, Centre 1 Station de Ge notypage, Evry, France, 3 UMR de Ge ne tique Ve ge tale, INRA, Universite Paris-Sud, CNRS, Institut National Agronomique Paris-Grignon, Gif-sur-Yvette, France, National de Ge orique et Mode `les Statistiques, Universite Paris-Sud, Orsay, France 4 Laboratoire de Physique The

In many species, sex-related differences in crossover (CO) rates have been described at chromosomal and regional levels. In this study, we determined the CO distribution along the entire Arabidopsis thaliana Chromosome 4 (18 Mb) in male and female meiosis, using high density genetic maps built on large backcross populations (44 markers, .1,300 plants). We observed dramatic differences between male and female map lengths that were calculated as 88 cM and 52 cM, respectively. This difference is remarkably parallel to that between the total synaptonemal complex lengths measured in male and female meiocytes by immunolabeling of ZYP1 (a component of the synaptonemal complex). Moreover, CO landscapes were clearly different: in particular, at both ends of the map, male CO rates were higher (up to 4-fold the mean value), whereas female CO rates were equal or even below the chromosomal average. This unique material gave us the opportunity to perform a detailed analysis of CO interference on Chromosome 4 in male and female meiosis. The number of COs per chromosome and the distances between them clearly departs from randomness. Strikingly, the interference level (measured by coincidence) varied significantly along the chromosome in male meiosis and was correlated to the physical distance between COs. The significance of this finding on the relevance of current CO interference models is discussed.
rard A, Falque M, et al. (2007) Sex-specific crossover distributions and variations in interference level along Arabidopsis Citation: Drouaud J, Mercier R, Chelysheva L, Be thaliana Chromosome 4. PLoS Genet 3(6): e106. doi:10.1371/journal.pgen.0030106

Introduction
One prominent feature of the eukaryotic life cycle is the segregation of homologous chromosomes to two different cells during the rst, also known as reductional, meiotic division. The proper completion of this segregation relies on the formation of stable physical connections between homologous chromosomes. In most eukaryotic species, these connections are mediated by crossovers (COs). These are sites where large (megabase scale) segments of homologous (nonsister) chromatids are exchanged. Consequently, COs are essential to the ploidy reduction process, as well as to play a role in the creation of allelic combinations. CO number and distribution along chromosomes differ between male and female meiosis in many plant and animal taxa (for review see [1]). This widespread phenomenon is called heterochiasmy. Both the direction and magnitude of these differences are highly variable. For example, depending on the species, CO number may be higher in female (F) meiosis (most eutherian mammals), or male (M) meiosis (some metatherian mammals), or there may be no signicant difference between sexes (goat, dog, barley). This difference may be small or moderate, but sometimes it is huge (e.g., teleostean shes). Even closely related species can exhibit different M/F CO ratios. In the Brassicaceae, for example, this ratio reaches 1.2 in Sinapis alba [2], whereas in Brassica oleracea it is inversed (0.6) [3], and there is no signicant difference in Brassica napus (0.98) [4]. Therefore, the nature of evolutionary forces driving heterochiasmy is a puzzling issue. In addition,
PLoS Genetics | www.plosgenetics.org 1096

the underlying molecular and cellular mechanisms are currently unknown. Yet sex-related differences in CO number per chromosome are paralleled by sex-related differences in the length of synaptonemal complex (SC) in human [5] and mouse [6]. The SC is a proteic structure scaffolded along synapsed homologous chromosomes at pachytene stage [7]. COs can be localized along chromosomes by analyzing genetic recombination data. They can also be visualized cytologically either as chiasma, or as immunolabeled MLH1 foci that mark most CO sites [8], or as late recombination nodules [9], which are electron-dense structures located on SCs [911]. COs originate from programmed double-strand breaks (DSBs) that occur early in prophase of the rst meiotic
Editor: Michael Lichten, National Cancer Institute, United States of America Received January 31, 2007; Accepted May 14, 2007; Published June 29, 2007 A previous version of this article appeared as an Early Online Release on May 15, 2007 (doi:10.1371/journal.pgen.0030106.eor). Copyright: 2007 Drouaud et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Abbreviations: CO, crossover; DSB, double-strand break; F, female; M, male; NCO, noncrossover; SC, synaptonemal complex * To whom correspondence should be addressed. E-mail: mezard@versailles.inra.fr [ These authors contributed equally to this work. ne tique Humaine, CNRS, Montpellier, France Current address: Institut de Ge

June 2007 | Volume 3 | Issue 6 | e106

65

Crossover Distribution in Arabidopsis

Author Summary
Meiotic crossovers between homologous chromosomes ensure their proper segregation to generate ultimately gametes. They also create new allelic combinations which contribute to the diversity of traits among individuals. In all eukaryotes, the number and the localization of crossovers along chromosomes are not random. In addition, crossovers are not independent of each other: the occurrence of a crossover lowers the probability that another crossover arises in its vicinity. The mechanism of this phenomenon, called crossover interference, is one of the most challenging puzzles that geneticists have been faced with in the last century. In this paper, we precisely described the distribution of crossovers along Chromosome 4 of the model plant species Arabidopsis thaliana, separately in male and female meiosis. Interestingly, we observed that crossovers are 1.7 more numerous in male than in female meiosis, and this increase is especially marked at the ends of the chromosome. Moreover, our results provide the first evidence that the level of interference along a chromosome is not a constant and is correlated with the physical distance between crossovers. These results shed new light on the determinism of crossover localization and could have important outcomes on the relevance of current models of crossover interference.

division [12]. Only a part of these DSBs give rise to COs; the remaining DSBs are repaired as noncrossovers (NCOs), without exchange of large DNA segments between homologous chromosomes. Numerous studies showed that CO formation is tightly controlled at both chromosomal and local scales [13,14]. Indeed, COs are not uniformly distributed and inter-CO distances are not random. The former feature is well illustrated by numerous datasets in mammals [1517] and higher plants [11,18]. Several studies have tried to correlate CO rates along chromosomes with various sequence features, such as gene or transposable element density, GC nucleotides %, CpG ratio, simple repeats, etc. However, even if some weak correlations were found, it seems that none holds true in all species [15,16,1921], suggesting that other constraints act on CO distribution. One of these constraints is CO interference. This phenomenon was originally described as a lower frequency of doubleCOs in disjoint chromosomal segments than expected if they occur independently of each other [22]. The existence of interference has been conrmed in most species tested [10]. As a consequence of interference, COs tend to be more evenly spaced than expected if CO positions were random [23]. In addition, in many species, which have a limited number of COs per chromosome, interference tends to increase physical distances between adjacent COs. This is well illustrated by recombination nodules or MLH1 foci maps produced in various species [5,17,2426]. The mechanisms of interference setup are still poorly understood. Several models of meiotic CO interference have been proposed over years (see [14] for a comprehensive review). The two main contenders are currently the counting model [27] and the mechanical stress model [28]. The basic postulate of the counting model is that the CO designation process among recombination precursors occurs in such a way that any two adjacent COs are separated by a xed number of NCOs. Alternatively, the mechanical stress model hypothesizes that COs originate from a mechanical
PLoS Genetics | www.plosgenetics.org 1097

stress imposed on the chromosome. CO designation would promote a stress relief that would (i) inhibit CO designation among nearby recombination intermediates and (ii) attenuate in a distance-dependent manner. Neither of these two models is presently strongly supported by experimental data. In a previous study, we produced a high resolution map (at around the 210-kb scale) of meiotic crossovers on Arabidopsis thaliana Chromosome 4 [18]. We showed that CO rates vary greatly along the chromosome from 0 to 20 cM/Mb, and that COs displayed interference. However, CO rates on this map were sex-averaged because we used the selfed progeny of F1 hybrids for the mapping population. Given that the existence of heterochiasmy in A. thaliana had been previously suggested by several studies [2932], we decided to investigate the relative contributions of male and female meiosis in the distribution of COs on Chromosome 4. We observed dramatic differences between male and female genetic maps. Strikingly, we found a good correlation between the sex-ratio of mean CO number per Chromosome 4 on one hand and the sex-ratio of total SC length on the other hand. Moreover, we were able to detect signicant variations in interference strength along Chromosome 4. Stunningly, it turned out that interference strength covaries with the physical distance between COs. These results could have important upshots on the reliability of current interference models.

Materials and Methods Generation of Backcross Populations and Genomic DNA Extraction
A. thaliana Columbia (Col) and Landsberg erecta (Ler) accessions were crossed to obtain F1 hybrids. Col plants were then crossed with an F1 hybrid used either as the male (Col 3 (Col 3 Ler)) or as the female ((Col 3 Ler) 3 Col) parent. Seeds from these crosses were sowed in vitro, and then seedlings were grown in short-day conditions at 21 8C. After 2 wk, 1,476 whole seedlings of each population were collected and their DNA was extracted as described previously [33].

Choice of Markers and Single Nucleotide Polymorphism Genotyping


In a previous experiment, F2 plants from a Col 3 Ler cross were genotyped with a set of 70 SNP markers spanning A. thaliana Chromosome 4 [18]. In the present study, 46 SNPs out of these 70 and two additional SNPs were chosen so that the mean sex-averaged genetic distance between adjacent markers was 1.9 cM. SNPs are listed in Table S1. Genotyping was performed using SNPlex technology (Applied Biosystems, http://www.appliedbiosystems.com) following the supplier protocols. After quality scoring of genotyping data, four markers were dismissed from the whole dataset. In some cases it was not possible to assess the genotype of remaining markers in some plants so these were also removed from the dataset. The resulting populations comprised 1,305 and 1,419 plants for female and male meiosis, respectively. The genetic size of intervals was computed as the ratio between the number of recombined chromosomes and the number of analyzed meioses, which in the case of a backcross progeny is equal to the number of analyzed plants. CO rates, physical and genetic sizes are listed in Table S2.
June 2007 | Volume 3 | Issue 6 | e106

66

Crossover Distribution in Arabidopsis

Comparisons of CO Rates
In order to calculate single-interval, sex-averaged CO rates in the pool of M and F populations, the sex-specic CO rates were weighted according to the respective population size. All pair-wise comparisons between CO rates were performed using a chi-square homogeneity test. For multiple testing, p-values were subsequently corrected using the false discovery rate procedure [34].

Comparisons of CO Number per Chromosome


Predicted Poisson distributions of CO number per chromosome were calculated using the following formula: m k Sk N e k!m where S(k) is the number of chromosomes harboring exactly k CO, e is the neperian logarithm base, m is the observed mean number of CO per chromosome, and N is the total number of chromosomes. Whole comparisons between observed and Poisson distributions were performed using a chi-square goodness-of-t test.

comparisons, for the same reasons (see above). Given two intervals, the coefcient of coincidence between them is 11 calculated as follows: C r10 r11r r01 r11 where C is the coincidence and rij is the chance of i CO across the rst interval and j CO across the second interval. In most cases i and j values are either 0 or 1; 2 COs were rarely found in one of the intervals and were considered as no CO, while 3 COs were considered as 1 only, accordingly to what would have been observed if the intervals would have not contained internal markers. The standard deviation of coincidence was calculated according to [35].

Testing for Significant Differences between Coincidence Values


We have developed a procedure which computes the pvalue for the hypothesis H0 that two coefcients of coincidence cu and cv estimated from quadruplets or triplets of markers are in fact generated from the same theoretical coincidence value cth. Under that hypothesis, H0, cu, and cv are actually not expected to differ from each other. A small pvalue for the difference between cu and cv then indicates that H0 is unlikely to be true given the genotype data of the mapping population. Let a quadruplet have markers A, B, C, and D, assumed to be in the order in which they appear on the chromosome. If the quadruplet is instead a triplet, this formalism can be applied by setting B C. In a rst phase, we compute cth by the maximum likelihood method. Consider the rst quadruplet: the probability (likelihood) that N gametes lead to a measured coincidence value of cu is N! rr nrn nnr nnn L cu p n rr prn pnr pnn nrr nrn nnr nnn where nnn, nrn, nnr, nrr are the number of gametes that are respectively recombinant between (i) neither A and B nor C and D, (ii) A and B but not C and D, (iii) C and D but not A and B, (iv) both A and B and C and D. N is the total number of gametes with valid data at the four markers, namely nnn nrn nnr nrr. The dependence on cth is through the probabilities: Prr rAB rCD cth Pnr 1 cth rAB rCD Prn rAB :1 cth rCD Pnn 1 rAB 1 rCD rAB :rCD 1 Cth where rAB and rCD are recombination fractions between A and B and B and C, respectively. Next, we consider the two quadruplets of interest. L(cv) is calculated as for L(cu). The joint likelihood of both observations is the product L(cu) 3 L(cv), and we numerically determine the cth which maximizes this joint likelihood. The result is a cth lying somewhere between cu and cv. In a second phase, we compute a p-value for the hypothesis H0 given cth. We do this by determining the probability that j cu cv j is at least as large as measured from the experimental data. But if cu and cv are estimated using shared gametes, the recombination events in the four intervals are a priori
1098 June 2007 | Volume 3 | Issue 6 | e106

Comparisons of Inter-CO Distances


For both M and F datasets, the genetic width of inter-CO distance classes was chosen to be 17.5 cM (6 5%) in order to: (i) provide distance classes spanning the whole chromosome genetic length, (ii) ensure a common denominator in both M and F maps, and (iii) prevent small class sizes, in order to maintain moderate sampling variances, thus allowing conclusive statistical testing. The continuous probability distribution function of interCO distances on chromosomes with exactly a independent da1 COs is: P d a LL , where L is the genetic size of the a chromosome and d is the distance between successive COs. The derivation of this formula is as follows: a independent CO points are randomly placed on a chromosome of length L. Then a CO point is added at one end to bring the chromosome to the shape of a ring. a 1 points are thus randomly and independently positioned on the circle of perimeter L. The statistics of distances between successive COs is the same for all pairs; to compute this for the rst pair, we need to nd the distribution of the smallest of a random variables, representing the positions of COs along the interval, which are uniformly distributed in [0, L] (the remaining point is by denition at position zero). The probability that this smallest value, which is the distance between the last and the rst CO, is greater or equal to X is (1 X/L)a. The minus derivative of this cumulated distribution then gives the desired probability distribution. The following formula, which is easily deduced from the formula above, allows convenient calculation of discrete distributions of inter-CO distances on nite-size chromok1N somes with exactly two independent COs: Sk 2nn , 2 where n is the number of classes, k is the rank of the class (increasing with distance), N is the population size, and S(k) is the size of the kth9 class. Whole comparisons between observed and calculated distributions were performed using a chi-square goodnessof-t test.

Coincidence Analyses
For both M and F datasets, the genetic size of intervals used for coincidence analyses was chosen to be the same as the genetic width of distance classes used for inter-CO distance
PLoS Genetics | www.plosgenetics.org

67

Crossover Distribution in Arabidopsis

Cytological Observations
Cytological observations were carried out on Col 3 Ler F1 plants. The anti-ASY1 polyclonal antibody has been described elsewhere [36]. It was used at a dilution of 1:500. The antiZYP1 polyclonal antibody was described by [37]. It was used at a dilution of 1:500. Preparation of prophase stage spreads for immunocytology was performed according to [36] with the modications described in [38]. All observations were made using a Leica (http://www.leica. com) DM RXA2 microscope; photographs were taken using a CoolSNAP HQ (Roper, http://www.roperscientic.com) camera driven by Open LAB 4.0.4 software; all images were further processed with Open LAB 4.0.4 or AdobePhotoshop 7.0 (http://www.adobe.com). SC length measurement was performed using Optimas (Bioscan Incorporated, http:// www.bioscan.com) software.

Results
Plants from a Col 3 Ler F1 population were backcrossed with Col plants using the F1, either as the male or the female parent, in order to create two populations subsequently referred to as M and F, in which the observed recombination events occurred either in male or female meiosis of the parental F1 hybrid. 1,419 M plants and 1,305 F plants were genotyped with 44 SNP markers spanning Chromosome 4 at a density of 1.9 cM (calculated from sex-averaged data, see Materials and Methods). Given that interval sizes are small, we calculated genetic distances simply by dividing the number of recombinant chromosomes by the total number of plants analyzed.
Figure 1. Variation of CO Rates along A. thaliana Chromosome 4 (A) CO rates in F2 population (green) and the pool of male and female backcross populations (red). (B) Alignment of physical map (center) and both male (top) and female (bottom) genetic maps. (C) CO rates in male (blue) and female (pink) populations. Orange stars mark the intervals that are significantly different between both populations (p-values , 0.02). A schematic representation of Chromosome 4 is aligned with the physical map and each CO plot, which includes 5-Mb scale coordinates, centromere (diamond), heterochromatic knob (gray box), nucleolar organizer region (NOR, black box). doi:10.1371/journal.pgen.0030106.g001

The CO Landscape on Chromosome 4 Does Not Differ between an F2 Population and Pooled Backcross Populations
We rst compared CO rates in the F2 population previously described to those in pooled M and F populations, in each of the same 43 intervals spanning Chromosome 4 (Figure 1A). As expected, the averaged (see Materials and Methods) CO rates observed in the pool of the M and F backcross progenies (corresponding respectively to male and female meiosis) were not signicantly different from those observed in the F2 progeny (resulting half from male meiosis, half from female meiosis) generated from the same parental accessions (lowest p-value is 0.35; Figure 1A). This implies that there is no signicant variation in meiotic recombination over time for a given genetic background, thus enabling direct comparisons of data.

correlated. Thus, when measuring cu and cv we need to use independent sets of gametes by using half (N/2) of the gametes for cu and the other half for cv. So that the value j cu cv j is not dependent on the data order, j cu cv j is computed for 105 random order combinations and the median value taken. The p-value is obtained by simulating interference events within H0 given cth: we generate N/2 realizations of gametes for each quadruplet; for each realization, we choose among the four possibilities of recombinants or not in each interval according to the probabilities prr, prn, pnr, pnn. For this set of N gametes, we extract the two associated coincidence coefcients cu9 and cv9. Repeating this 105 times, we get a probability distribution for j cu9 cv9 j; the desired p-value is then the frequency with which j cu9 cv9 j is larger than the experimental value.
PLoS Genetics | www.plosgenetics.org 1099

The CO Landscape on Chromosome 4 Differs in Male and Female Meiosis


At rst glance, the difference between male and female recombination rates is obvious when comparing total genetic size of both maps (Figure 1B). The M map is 87.9 cM long and the F map is 52.3 cM long. This indicates that a Chromosome 4 bivalent experiences on average 1.76 CO in male meiosis, but only 1.05 CO in female meiosis (M/F ratio 1.68). This M/F difference is highly signicant (v2 p-value , 0.001) Next, we compared recombination rates in male and female meiosis interval-by-interval (Figure 1C). For a majority
June 2007 | Volume 3 | Issue 6 | e106

68

Crossover Distribution in Arabidopsis

Figure 2. Coimmunolocalization of ASY1 (Red) and ZYP1 (Green) in Male and Female Meiocytes at Pachytene Stage Bar 10 lm. doi:10.1371/journal.pgen.0030106.g002

of intervals (36/43) the M/F ratio was above 1, with the most notable differences in the last telomeric third of the long arm. However, only the distal interval on the short arm and the ve distal intervals on the long arm were highly signicantly different in male and female (mean M/F ratio for these six intervals is 6.1, v2 p-value , 0.05). The remaining central intervals were not signicantly different in male and female, when compared one-by-one. However, if these were grouped and considered as a single interval, there was still a signicant difference between male and female (M/F ratio 1.37, v2 pvalue , 0.001). In summary, male and female meiotic CO landscapes along Chromosome 4 are strikingly different. The difference is high close to the telomere on the long arm and to the nucleolar organizer region on the short arm and modest in the median region of the chromosome (see Figure 1B and 1C).

Figure 3. Distributions of CO Number per Chromosome 4 Bars represent the frequency of chromatids with 0, 1, 2, or 3 COs. (A) Observed (blue) and Poisson (purple) distributions in M population. (B) Observed (blue) and Poisson (purple) distributions in F population. Corresponding number of chromatids is indicated above each bar. doi:10.1371/journal.pgen.0030106.g003

Total SC Length Differs in Male and Female Meiosis


Meiotic chromosomes at pachytene stage were immunolabeled with antibodies against ZYP1. This protein is a major component of the central element of the SC, which ties homologous chromosomes together. We used ASY1 immunolabeling to visualize the axial element, which is a proteinaceous axis formed along pairs of sister chromatids [39] (Figure 2). At pachytene stage, ZYP1 labeling extends continuously along the entire SC, hence allowing total SC length measurement. We found that SC length in male meiocytes is 166 6 24 lm (n 22) compared to only 98 6 20 lm (n 25) in female. Our estimate of male SC length is in good agreement with that obtained in a previous study (147 6 28 lm; n 19) using electron microscopy [40]. The value we obtained for the M/F ratio of total SC lengths is very close to that for the M/F ratio of mean CO numbers per chromosome (1.70 versus 1.76). This suggests that sex-related differences in CO number and total SC length are correlated.

and independent events, the distribution of CO number per chromosome should t a Poisson distribution. Thus, we calculated the Poisson distributions expected for the observed average number of COs per Chromosome 4 and compared these to the observed ones (see Figure 3A and 3B). In the F population, about half of chromosomes had no CO or only one CO, while very few had two COs or more (Figure 3B). This distribution is highly signicantly different (p-value , 0.001) from the theoretical Poisson distribution, in which the 0 CO group was the main class (59%) and multiple CO classes accounted for 10%. Hence, in female meiosis almost all bivalents experienced only the obligate CO required for the proper segregation of homologous chromosomes at anaphase I. In the M population, only one third of chromosomes had no CO and about half had one CO (Figure 3A). Consequently, chromosomes with multiple COs were more frequent than in the F population. Conversely, in the corresponding Poisson distribution 0 CO and 1 CO chromosomes were represented at 42% and 36%, respectively. Observed and expected distributions were clearly different from each other (p-value , 0.001).

Distributions of CO Number per Chromosome in Male and Female Meiosis Do Not Fit Poisson Distributions.
We next looked at the distribution of CO number per Chromosome 4 in the M and F populations (Figure 3). According to the hypothesis that CO placements are random
PLoS Genetics | www.plosgenetics.org 1100

Distances between Adjacent COs Are Greater than Expected


As a consequence of interference, inter-CO distances are less variable and greater (when CO number is limited) than expected under the assumption that COs are distributed
June 2007 | Volume 3 | Issue 6 | e106

69

Crossover Distribution in Arabidopsis

Figure 5. Distributions of Inter-CO Distances on Chromosome 4 Observed (blue) and random (purple) distributions of inter-CO distances in (A) F and (B) M populations. doi:10.1371/journal.pgen.0030106.g005 Figure 4. Double-CO Locations on (A) Male and (B) Female Chromosome 4 All double-COs on chromosomes harboring exactly two COs were plotted in two-dimensional graphs. The x and y axis values indicate the genetic position (relative to the nucleolar organizing region) of the first and the second CO, respectively. The diagonal line and the upper left corner correspond to minimal (null) and maximal (chromosome-wide) inter-COs distances, respectively. doi:10.1371/journal.pgen.0030106.g004

randomly and independently. Positions of double-COs (on chromosomes with exactly two COs) were represented in twodimensional plots in Figure 4. x and y axis coordinates correspond to the positions of the rst and second CO on the genetic map, respectively. Under the assumption of no interference, points should be uniformly distributed over the triangle. For both M and F double-CO populations, the observed points were clearly heterogeneously distributed: they were underrepresented next to the diagonal line, which corresponds to low inter-CO distances. In order to test this deviation from independence between COs, inter-CO distances on chromosomes with two COs only (see Materials and Methods) were grouped into size classes, and the observed distributions were compared to the random (no CO interference) distributions (Figure 5). For both M and F datasets, the genetic length (17.5 cM 6 5%) of the intervals was chosen to optimize the number of doublePLoS Genetics | www.plosgenetics.org 1101

COs per interval, in order to avoid high sampling variance and thus allow statistically signicant differences to be detected. In the F population, we found opposing observed and expected distributions: for the expected distribution the minor class was 3552.5 cM and the major class 017.5 cM, whereas in the observed distribution the majority of inter-CO distances were long, and short distances were the minority (Figure 5A). This difference was highly signicant (p-value , 0.001). The observed M distribution was rather symmetrical, with the mode between 35 and 53 cM. It was strikingly different from the theoretical distribution, in which the class size decreased with increasing genetic length (p-value , 0.001; Figure 5B). For both M and F distributions the mean observed interCO distance, respectively 51% and 63% of the total map size, exceeded the expected one, which is exactly one third of the total map size. Therefore, in male and female meiosis, widely spaced COs were overrepresented, whereas closely spaced COs were underrepresented. This difference between expected and observed distributions of distances between COs is fully consistent with interference.
June 2007 | Volume 3 | Issue 6 | e106

70

Crossover Distribution in Arabidopsis

where rij is the chance of i CO across the rst interval and j CO across the second interval. The value of C is 1 if there is no interference and 0 if interference is absolute (meaning that double-COs are completely absent). Coincidence is widely used as a measure of interference from genetic data. Moreover, most mathematical models of CO interference assume a covariation between coincidence at a given genetic distance and the level of interference (see for example [27], reviewed in [14] ). Hence, plotting coincidence for pairs of adjacent intervals (three-point coincidence: C3; [27]) all along a chromosome gives access to local variations of interference level, provided that the genetic size of intervals remains constant. We thus performed all coincidence analyses on every possible pairs of 17.5 cM (6 5%) adjacent intervals (30 and 21 pairs t these requirements in M and F datasets, respectively). This means that we moved a 2 3 17.5-cM window along the genetic map. The interference level measured by C3 was clearly variable across Chromosome 4 for both maps. In male meiosis, starting from the short-arm end, interference strength was high until ;30 cM (C3 , 0.1), then it decreased from ;30 cM to ;45 cM (C3 ;0.3), to reach a minimum at ;52 cM (C3 ;0.75), and nally increased again from ;65 cM to the end of the map (C3 ;0.3; Figure 6A). Most of these variations in C3 were found to be signicant (see Figure 6, Table 1, and Materials and Methods). We can thus conclude that local interference level varied signicantly along Chromosome 4 in male meiosis. In the F plot, all observed C3 values are very low ( 0.1). We could not observe any signicant variation in coincidence among the few points of the plot (Figure 6 and Table 1). Given the very small number of double-COs, it seems likely that many more plants would be needed to detect any possible coincidence variation.

Four-Point Coefficient of Coincidence Increases with Genetic Distance


Figure 6. C3 Coincidence Plots along the (A) Male or (B) Female Chromosome 4 All possible pairs of adjacent intervals (each being 17.5 cM 6 5%) were used for calculation of coefficient of coincidence, defined as the frequency of double-COs divided by the frequencies of COs in both intervals (see Materials and Methods). These C3 coincidence values were plotted against the genetic position of the junction between the two adjacent intervals. Blue and pink dots represent, respectively, M and F datasets. Numbered green dots mark the pairs of intervals used for statistical comparison of C3 coincidence values. Error bar for these dots correspond to the standard deviation of coincidence calculated according to [35]. Each plot is aligned with the corresponding genetic map, the physical map, and a schematic representation of the chromosome, which includes 5-Mb scale coordinates, centromere (diamond), heterochromatic knob (gray box), nucleolar organizer region (NOR, black box). doi:10.1371/journal.pgen.0030106.g006

Three-Point Coefficient of Coincidence Varies along Chromosome 4


Besides altered inter-CO distances, another expected consequence of interference is a lowered chance of nding close double-COs than expected from randomness. More precisely speaking, given two intervals, double-COs (one CO in each interval) will occur at a lower frequency than two independent COs (one CO in the rst or in the second interval, both being not exclusive). This departure is called 11 coincidence and can be calculated as follows: c r10 r11r r01 r11
PLoS Genetics | www.plosgenetics.org 1102

Another way of analyzing interference is to plot coincidence between one xed interval and a series of increasingly distant intervals (four-point coincidence: C4; [27]). This means that we xed a 17.5 cM (6 5%) window and moved a second 17.5 cM (6 5%) window all along the chromosome. This method provides a global description of interference depending on genetic distance (see Materials and Methods). For each M and F population, two different C4 plots were made, using either the terminal interval of the short arm (Figures 7A and 8A) or the telomeric interval of the long arm as the xed interval (Figures 7B and 8B). For the M population, regardless of whether the xed interval was located at the end of the long or short arm, plots had globally the same shape. This kind of shape has been consistently observed in various species (for review see [41]), showing that interference decreases with genetic distance. Nevertheless, C4 values from the short-arm end were systematically lower (from 0.05 to 1.11) than those from the long-arm end (from 0.3 to 1.23). This conrms that the strength of interference is weaker at the distal region of the long arm than on the short arm. In both plots, coincidence increased up to 1 at ;4550 cM, peaked above 1 at ;5560 cM, and then decreased toward 1. The shape of the C4 plots was determined by two factors: (i) the genetic distance between intervals and (ii) uctuations of interference strength as described above. The occurrence of a peak of coincidence above 1 means that at
June 2007 | Volume 3 | Issue 6 | e106

71

Crossover Distribution in Arabidopsis

Table 1. Comparison between C3 Values Obtained at Different Positions along Chromosome 4 in M and F Meiosis
M Population
Point number (C3 value) 1 (0.047) 2 (0.179) 3 (0.755) 2(0.179) 0.185 3(0.755) ,104 0.002 4(0.321) 0.034 0.328 0.039

F Population
Point number (C3 value) 1 (0.072) 2(0.096) 0.540

C3 values at evenly spaced positions (numbered from 1 to 4 on the M plot, 1 and 2 on the F plot) along each plot (green dots in Figure 6) were compared using a simulation-based approach (see Materials and Methods). For each pair-wise comparison, p-values (in red) given in the table correspond to the H0 hypothesis that the intensity of interference is actually the same at both positions. For each point (in bold), C3 value is indicated between parentheses. doi:10.1371/journal.pgen.0030106.t001

some genetic distances, double-COs are more frequent than expected if COs were randomly placed: this corresponds to what is called negative interference. At short distances from the terminal short-arm interval (which includes the centromere) coincidence was very low. Interestingly, this shows that the presence of the centromere did not block interference. Regarding the F population (Figure 8), because the F map was very short (52.4 cM) and double-COs were rare, C4 plots were less informative. From both ends, coincidence increased up to ;0.4 at ;35 cM. Strikingly, C never reached 1, showing that in female meiosis interference acted across the whole chromosome.

Discussion
In this study, we compared male and female meiotic CO distributions along A. thaliana Chromosome 4. We depicted strong sex-related differences in CO distribution (heterochiasmy) and unequivocal variations in CO interference level along the male chromosome. One major nding in this study was that the genetic sizes of male and female maps were strikingly different (M map 88 cM, F map 52 cM). This means that Chromosome 4 harbors an average of 1.8 CO in male meiosis, but only 1 in female meiosis. Noticeably, the size of our male map was very close to that previously reported for the Col accession [42] (87.9 versus 83.6 cM). Moreover, we described a marked difference in total SC length between sexes. Remarkably, this difference was very close to the M/F CO ratio observed for Chromosome 4. By using a high density of markers, as well as a large-sized population, we could describe the sex-specic ne scale distribution of COs for the rst time in A. thaliana. Thus, we could detect highly signicant variations of M/F CO ratio along the chromosome. We observed large differences (up to 18.7-fold) at both ends of the genetic map and less pronouncedthough signicantdifferences on the median part of the chromosome (from 5 Mb to 13 Mb, see Figure 1C). In this median region both curves presented the same overall pattern of peaks and valleys, even if the local male CO rate was higher than the female CO rate for most intervals (30/37). In summary, we detected heterochiasmy, not only at the whole chromosome level, but also at a regional scale. Heterochiasmy in A. thaliana has been suggested by several previous studies. Indeed, chiasma counts showed that COs are more frequent in male meiosis (9.7) than in female (8.5) [32], but the result of this count, carried out on only ten meiocytes, was not highly signicant and did not provide any precise information about CO location. In two other studies, a comparison of male and female recombination rates on all ve A. thaliana chromosomes was performed [29,31], but in each case only a few markers were scored, covering only part of the genome. Nevertheless, these results also suggested that recombination is higher in male than female meiosis and that there are variations in the M/F CO ratio along chromosomes. At the whole genome scale, heterochiasmy is widespread (reviewed in [1]), but this observation remains largely unexplained from an evolutionary point of view. Nevertheless, SC length was shown to differ signicantly between male and female meiocytes in human [5] and mouse [6], paralleling variations in MLH1 foci number. Similar results
1103 June 2007 | Volume 3 | Issue 6 | e106

Level of Interference Covaries with Physical Size


Given that signicant variations of interference level along Chromosome 4 were detected in male meiosis, we addressed the issue of possible correlations between interference level and physical distance. We thus calculated the physical size of the pairs of intervals considered for the C3 coincidence analysis described above, which have all the same genetic size (2 3 17.5 cM). The coincidence values were next plotted against these sizes (see Figure 9A). We could clearly observe two scatters of points. The smallest one contains all pairs of intervals encompassing the heterochromatic knob located on the small arm and the centromere, while the largest scatter comprises all the pairs of intervals located on the long arm only. For the large scatter, we could note a striking positive correlation between the C3 coincidence and the physical size (r2 0.91 6 0.04). The small scatter was shifted by about 4.5 Mb relative to the large scatter. This is a direct consequence of the presence of a large CO-free region, which does not contribute to the genetic size of the considered pairs of intervals, but hugely increases their physical size. Indeed, when subtracting the cumulated size of the knob and the centromere (4.7 Mb) from the size of the pairs of intervals encompassing these chromosome parts and making a new plot (see Figure 9B), the two scatters of Figure 9A grouped into a single scatter that showed this time a chromosome-wide tight correlation between C3 coincidence and physical size (r2 0.93 6 0.04). It means that interference levelmeasured by coincidence analysis on adjacent intervals of constant genetic size decreases as the physical size increases. In other words, for a given genetic size (i.e., a given CO frequency), a greater physical size enhances the opportunity for double-COs to occur.
PLoS Genetics | www.plosgenetics.org

72

Crossover Distribution in Arabidopsis

Figure 7. C4 Coincidence Plots along the Male Chromosome 4 All possible combinations between a fixed and a nonoverlapping interval (each being 17.5 cM 6 5%) were used for calculation of coefficient of coincidence, defined as the frequency of double-COs divided by the frequencies of COs in both intervals (see Materials and Methods). These C4 coincidence values were plotted against the genetic distance between the centers of the intervals. The fixed interval is located at the end of either (A) the short arm or (B) the long arm. Each plot is aligned with the corresponding genetic map, the physical map, and a schematic representation of the chromosome, which includes 5-Mb scale coordinates, centromere (diamond), heterochromatic knob (gray box), nucleolar organizer region (NOR, black box). For both plots, abscissa values indicate the genetic distance between the centers of the fixed and the moving intervals. The red two-headed arrow indicates the fixed interval position. doi:10.1371/journal.pgen.0030106.g007

Figure 8. C4 Coincidence Plots along the Female Chromosome 4 All possible combinations between a fixed and a nonoverlapping interval (each being 17.5 cM 6 5%) were used for calculation of coefficient of coincidence, defined as the frequency of double-COs divided by the frequencies of COs in both intervals (see Materials and Methods). These C4 coincidence values were plotted against the genetic distance between the centers of the intervals. The fixed interval is located at the end of either (A) the short arm or (B) the long arm. Each plot is aligned with the corresponding genetic map, the physical map, and a schematic representation of the chromosome, which includes 5-Mb scale coordinates, centromere (diamond), heterochromatic knob (gray box), nucleolar organizer region (NOR, black box). For both plots, abscissa values indicate the genetic distance between the centers of the fixed and the moving interval. The red two-headed arrow indicates the fixed interval position. doi:10.1371/journal.pgen.0030106.g008

were reported in other species, such as zebrash [43,44], Dendrocoelum lacteum [45], and others (reviewed in [46]). This statement can now be extended to higher plants, since we obtained comparable results in A. thaliana. Such correlated variations in SC length and CO number were also reported among chromosomes in one sex only, among individuals in the same species, and even among meiocytes in a single
PLoS Genetics | www.plosgenetics.org 1104

organism [5,17,4750]. However, based on current knowledge it is difcult to claim that SC length determines CO number, if the reverse is true, or even if another unidentied factor determines both. Besides global differences, there is also compelling evidence that the distribution of COs along chromosomes is contrasted in both sexes. Similarly to what we observe in A.
June 2007 | Volume 3 | Issue 6 | e106

73

Crossover Distribution in Arabidopsis

Figure 9. Scatter Plot between C3 Coincidence Value and Physical Size for Male Chromosome 4 Violet dots correspond to pairs of intervals encompassing the heterochromatic knob and the centromere, while orange dots correspond to the remaining pairs of intervals (all located on the long arm). Numbered green dots represent the same pairs of intervals than those represented in Figure 6A and used for the comparisons of C3 coincidence values. (A) Physical sizes were calculated from genome sequence data (Build 6 version 0), adding if necessary 1.5 Mb to take account for the centromere. (B) Cumulated centromere and knob size (4.7 Mb) was subtracted from the size of the relevant pairs of intervals. doi:10.1371/journal.pgen.0030106.g009

thaliana, enhancement of the M/F ratio close to telomeres was reported in other Brassicaceae species [2,3]. In vertebrates, such a conservation in heterochiasmy patterns along chromosomes was also observed: in mouse, human, and several teleostean shes, the M/F ratio decreases around the centromeres and tends to increase close to the telomeres [15,16,43,51]. At the present time, the molecular and cellular bases of regional heterochiasmy remain elusive. And, generally speaking, the mechanisms ruling CO distribution along chromosomes are also poorly characterized. Data from various model organisms show that CO distribution results from the integration of several levels of control [14]: (i) the density of meiotic DSBs initiating recombination between homologous chromosomes varies along chromosomes; (ii) the propensity of a DSB to be repaired as a CO or a NCO probably varies; (iii) interference shapes the nal CO distribution; (iv) only a part (variable among species) of COs are sensitive to interference (type I CO), the remaining are insensitive (type II CO). Each of these layers of control could act on observed heterochiasmy patterns along chromosomes. The DSB distribution, CO/NCO ratio, interference strength/interference strength variations, and the proportion of type II COs could
PLoS Genetics | www.plosgenetics.org 1105

vary between male and female meiosis, even if no experimental data presently support these hypotheses. Other factors were also suggested to effect differences between CO distributions in male and female meiosis. Several studies suggested that in human, parental imprinting in a few regions could explain at least part of local heterochiasmy [52,53]. Additionally, it was proposed that synapsis initiation sites colocalize with COs [54]. For example, in human, synapsis initiation occurs in subtelomeric regions in male [55] whereas it is rather interstitial in female (reviewed in [56]), which seems compatible with the observed pattern of heterochiasmy. Due to the availability of whole genome sequences, correlations between CO rates along chromosomes and various genomic features could be examined [15,16,1921], but all the resulting correlations were weak. This could be explained by the fact that only sexaveraged recombination rates were used in these studies. A priori, there is no evidence that genomic features correlated to CO rates have the same weight in male and female meiosis. Thus, when possible, correlation analyses should be done separately on data from both sexes. Presumably, this could disclose previously unidentied relationships or reinforce existing ones and reveal differences in correlations between sexes. Altogether, it is likely that multiple constraints act synergistically to shape CO distribution along male and female chromosomes in meiosis, some of which remain to be elucidated. In this paper, we have presented the rst detailed study on the effect of interference on CO distribution along a whole chromosome in male and female meiosis of A. thaliana. Both CO number per chromosome and inter-CO distances clearly show that COs are not independent of each other. Interestingly, we unequivocally show that the centromere is not a barrier to interference, in accordance with previous reports [18,5759]. The coincidence plots also clearly show the existence of negative interference at some genetic distance (5560 cM), which corresponds to a greater chance of another CO than expected from random. This phenomenon has been repeatedly observed from various genetic datasets (see, for instance, [58]) and is also predicted by various models of CO interference [27,28]. Furthermore, we provide unambiguous evidence that interference strength varies signicantly along A. thaliana Chromosome 4 in male meiosis. It has recently been shown that in most eukaryotes, a part of meiotic COs arising from a distinct pathway are not sensitive to interference [60]. Such COs account for about 15% of the total in A. thaliana [37,61]. Thus, the observed variation of interference level, measured on the whole population of COs, can be explained by two nonexclusive hypotheses: (i) the interference level between interfering COs is actually variable, or (ii) this interference level is constant, but the relative proportions of the two kinds of COs are variable along chromosome, so that locally, a high density of noninterfering COs leads to a decrease of the interference level that is measured on the whole population of COs. In female meiosis, observed variations are not signicant because double-COs are rare, hence sampling variance is high, causing an increase in p-values from statistical testing. Such variations in interference strength along chromosomes were previously suggested from analysis of human pedigree data [62]. Moreover, the level of interference was
June 2007 | Volume 3 | Issue 6 | e106

74

Crossover Distribution in Arabidopsis

reported to vary among chromosomes in humans in several studies [17,6264]. Sex-linked variations of interference have also been reported along human Chromosome 21 [64]. Even uctuations of interference level among human individuals have been described [17,26]. The molecular bases of these variations are currently poorly documented. Our results provide clear evidence that across a chromosome segment displaying a given CO frequency, a greater physical size enhances the opportunity for double-COs to occur. In other words, interference level between COs separated by a xed genetic distance is a function of physical distance. Interestingly, cytogenetic data collected in humans demonstrate a negative correlation among chromosomes between SC length and the global (chromosomal) level of interference [17]. At the present time, the molecular bases of these variations are totally unknown. The mechanisms of interference itself are still elusive. Several models have been proposed, but no experimental data directly support them. One of the most widely used is the counting model. Its basic postulate is that a xed number of NCOs occurs between any two adjacent COs. As a consequence, interference strength is supposed to be constant at the chromosome scale [27,65]. Our data and those cited above strongly argue against such constancy and also call into question the concept of an unchanging count itself. Moreover, a recent study in yeast showed that CO number is maintained at the expense of NCOs when the DSB number is reduced, without affecting interference [66]. Other models propose that an interference signal results either from the progressive polymerization of a hypothetical structure along the chromosome [67], or from a mechanical stress imposed on the chromosome axis [28]. Our data are compatible with these two models, in which the interference
References 1. Lenormand T, Dutheil J (2005) Recombination difference between sexes: A role for haploid selection. PLoS Biol 3: e63. doi.10.1371/journal.pbio. 0030063 2. Nelson MN, Nixon J, Lydiate DJ (2005) Genome-wide analysis of the frequency and distribution of crossovers at male and female meiosis in Sinapis alba L. (white mustard). Theor Appl Genet 111: 3143. 3. Kearsey MJ, Ramsay LD, Jennings DE, Lydiate DJ, Bohuon EJR, et al. (1995) Higher recombination frequencies in female compared to male meisoses in Brassica oleracea. Theor Appl Genet 92: 363367. 4. Kelly AL, Sharpe AG, Nixon JH, Evans EJ, Lydiate DJ (1997) Indistinguishable patterns of recombination resulting from mate and female meioses in Brassica napus (oilseed rape). Genome 490: 4956. 5. Tease C, Hulten MA (2004) Inter-sex variation in synaptonemal complex lengths largely determine the different recombination rates in male and female germ cells. Cytogenet Genome Res 107: 208215. 6. Lynn A, Koehler KE, Judis L, Chan ER, Cherry JP, et al. (2002) Covariation of synaptonemal complex length and mammalian meiotic exchange rates. Science 296: 22222225. 7. Page SL, Hawley RS (2004) The genetics and molecular biology of the synaptonemal complex. Annu Rev Cell Dev Biol 20: 525558. 8. Anderson LK, Reeves A, Webb LM, Ashley T (1999) Distribution of crossing over on mouse synaptonemal complexes using immunouorescent localization of MLH1 protein. Genetics 151: 15691579. 9. Carpenter AT (1979) Synaptonemal complex and recombination nodules in wild-type Drosophila melanogaster females. Genetics 92: 511541. 10. Zickler D, Kleckner N (1999) Meiotic chromosomes: Integrating structure and function. Annu Rev Genet 33: 603754. 11. Anderson LK, Stack SM (2005) Recombination nodules in plants. Cytogenet Genome Res 109: 198204. 12. Keeney S (2001) Mechanism and control of meiotic recombination initiation. Curr Top Dev Biol 52: 153. 13. de Massy B (2003) Distribution of meiotic recombination sites. Trends Genet 19: 514522. 14. Mezard C, Vignard J, Drouaud J, Mercier R (2007) The road to crossovers: Plants have their say. Trends Genet 23: 9199.

level is not explicitly intended to be constant and an interference signal propagates along the chromosomes. However, as data in the eld of meiotic recombination continue to accumulate exponentially, it is likely that new CO interference models supported by experimental evidences will emerge in the near future. Our study provides the rst detailed analysis of heterochiasmy and CO interference at the whole chromosome scale in a plant species. It provides the basis for future investigations on the determinism of CO distribution at the whole genome scale in A. thaliana and other species.

Supporting Information
Table S1. List of SNP Markers Used for Genotyping Including Physical Position, 20-bp Left Flanking Sequence, and Alleles Found at doi:10.1371/journal.pgen.0030106.st001 (60 KB DOC). Table S2. Physical and Genetic Size of Intervals in Male and Female Populations Found at doi:10.1371/journal.pgen.0030106.st002 (64 KB DOC).

Acknowledgments
We are grateful to Mathilde Grelon and Eric Jenczewski for critical reading of the manuscript and Martine Leon for providing formula of inter-CO distances without interference. We thank the reviewers of this manuscript for their insightful suggestions. All the members of the Me iose et Recombinaison group provided helpful comments and participated in stimulating discussions. Author contributions. CM conceived and designed the experiments. JD, LC, AB, VZ, and CM performed the experiments. JD, RM, LC, MF, OM, and CM analyzed the data. JD, VZ, DB, and CM contributed reagents/materials/analysis tools. JD, RM, and CM wrote the paper. Funding. The authors received no specic funding for this study. Competing interests. The authors have declared that no competing interests exist.
15. Kong A, Gudbjartsson DF, Sainz J, Jonsdottir GM, Gudjonsson SA, et al. (2002) A high-resolution recombination map of the human genome. Nat Genet 31: 241247. 16. Shifman S, Bell JT, Copley RR, Taylor MS, Williams RW, et al. (2006) A highresolution single nucleotide polymorphism genetic map of the mouse genome. PLoS Biol 4: e395. doi.10.1371/journal.pbio.0040395 17. Codina-Pascual M, Campillo M, Kraus J, Speicher MR, Egozcue J, et al. (2006) Crossover frequency and synaptonemal complex length: Their variability and effects on human male meiosis. Mol Hum Reprod 12: 123 133. 18. Drouaud J, Camilleri C, Bourguignon PY, Canaguier A, Berard A, et al. (2006) Variation in crossing-over rates across Chromosome 4 of Arabidopsis thaliana reveals the presence of meiotic recombination hot spots. Genome Res 16: 106114. 19. Jensen-Seaman MI, Furey TS, Payseur BA, Lu Y, Roskin KM, et al. (2004) Comparative recombination rates in the rat, mouse, and human genomes. Genome Res 14: 528538. 20. Mezard C (2006) Meiotic recombination hot spots in plants. Biochem Soc Trans 34: 531534. 21. Gaut BS, Wright SI, Rizzon C, Dvorak J, Anderson LK (2007) Recombination: An underappreciated factor in the evolution of plant genomes. Nat Rev Genet 8: 7784. 22. Muller HJ (1916) The mechanism of crossing-over. Am Nat 50: 193434. 23. de Boer E, Stam P, Dietrich AJ, Pastink A, Heyting C (2006) Two levels of interference in mouse meiotic recombination. Proc Natl Acad Sci U S A 103: 96079612. 24. Sherman JD, Stack SM (1995) Two-dimensional spreads of synaptonemal complexes from solanaceous plants. VI. High-resolution recombination nodule map for tomato (Lycopersicon esculentum). Genetics 141: 683708. 25. Moens PB (2006) Zebrash: Chiasmata and interference. Genome 49: 205 208. 26. Sun F, Oliver-Bonet M, Liehr T, Starke H, Turek P, et al. (2006) Variation in MLH1 distribution in recombination maps for individual chromosomes from human males. Hum Mol Genet 15: 23762391. 27. Foss E, Lande R, Stahl FW, Steinberg CM (1993) Chiasma interference as a function of genetic distance. Genetics 133: 681691.

PLoS Genetics | www.plosgenetics.org

1106

June 2007 | Volume 3 | Issue 6 | e106

75

Crossover Distribution in Arabidopsis


28. Kleckner N, Zickler D, Jones GH, Dekker J, Padmore R, et al. (2004) A mechanical basis for chromosome function. Proc Natl Acad Sci U S A 101: 1259212597. 29. Vizir I, Korol A (1990) Sex difference in recombination frequency in Arabidopsis. Heredity 65: 379383. 30. Copenhaver GP, Browne WE, Preuss D (1998) Assaying genome-wide recombination and centromere functions with Arabidopsis tetrads. Proc Natl Acad Sci U S A 95: 247252. 31. Barth S, Melchinger AE, Devezi-Savula B, Lubberstedt T (2000) A highthroughput system for genome-wide measurement of genetic recombination in Arabidopsis thaliana based on transgenic markers. Funct Integr Genomics 1: 200206. 32. Armstrong SJ, Jones GH (2001) Female meiosis in wild-type Arabidopsis thaliana and in two meiotic mutants. Sex Plant Reprod 13: 177183. 33. Loudet O, Chaillou S, Camilleri C, Bouchez D, Daniel-Vedele F (2002) Bay-0 x Shahdara recombinant inbred line population: A powerful tool for the genetic dissection of complex traits in Arabidopsis. Theor Appl Genet 104: 11731184. 34. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: A practical and powerful approach to multiple testing. J R Stat Soc, Ser B, Methodol 57: 289300. 35. Muller HJ, Jacobs-Muller JM (1925) The standard errors of chromosome distances and coincidence. Genetics 10: 509524. 36. Armstrong SJ, Caryl AP, Jones GH, Franklin FC (2002) Asy1, a protein required for meiotic chromosome synapsis, localizes to axis-associated chromatin in Arabidopsis and Brassica. J Cell Sci 115: 36453655. 37. Higgins JD, Armstrong SJ, Franklin FC, Jones GH (2004) The Arabidopsis MutS homolog AtMSH4 functions at an early step in recombination: Evidence for two classes of recombination in Arabidopsis. Genes Dev 18: 25572570. 38. Chelysheva L, Diallo S, Vezon D, Gendrot G, Vrielynck N, et al. (2005) AtREC8 and AtSCC3 are essential to the monopolar orientation of the kinetochores during meiosis. J Cell Sci 118: 46214632. 39. Higgins JD, Sanchez-Moran E, Armstrong SJ, Jones GH, Franklin FC (2005) The Arabidopsis synaptonemal complex protein ZYP1 is required for chromosome synapsis and normal delity of crossing over. Genes Dev 19: 24882500. 40. Albini SM (1994) A karyotype of the Arabidopsis thaliana genome derived from synaptonemal complex analysis at prophase I of meiosis. Plant Journal 5: 665672. 41. McPeek MS, Speed TP (1995) Modeling interference in genetic recombination. Genetics 139: 10311044. 42. Lam SY, Horn SR, Radford SJ, Housworth EA, Stahl FW, et al. (2005) Crossover interference on nucleolus organizing region-bearing chromosomes in Arabidopsis. Genetics 170: 807812. 43. Singer A, Perlman H, Yan Y, Walker C, Corley-Smith G, et al. (2002) Sexspecic recombination rates in zebrash (Danio rerio). Genetics 160: 649 657. 44. Wallace BM, Wallace H (2003) Synaptonemal complex karyotype of zebrash. Heredity 90: 136140. 45. Croft JA, Jones GH (1989) Chromosome pairing and chiasma formation in spermatocytes and oocytes of Dendrocoelum lacteum (Turbellaria, Tricladida): A cytogenetical and ultrastructural study. Heredity 63: 97106. 46. Kleckner N, Storlazzi A, Zickler D (2003) Coordinate variation in meiotic pachytene SC length and total crossover/chiasma frequency under conditions of constant DNA length. Trends Genet 19: 623628. 47. Quevedo C, Del Cerro AL, Santos JL, Jones GH (1997) Correlated variation of chiasma frequency and synaptonemal complex length in Locusta migratoria. Heredity 78: 515519. 48. Zhang L, Murray BG, Pickering RA (2002) Variable patterns of chromosome synapsis at pachytene in Hordeum vulgare x H. bulbosum hybrids and their parents. Hereditas 137: 9095. 49. Froenicke L, Anderson LK, Wienberg J, Ashley T (2002) Male mouse recombination maps for each autosome identied by chromosome painting. Am J Hum Genet 71: 13531368. 50. Sun F, Oliver-Bonet M, Liehr T, Starke H, Ko E, et al. (2004) Human male recombination maps for individual chromosomes. Am J Hum Genet 74: 521531. 51. Sakamoto T, Danzmann RG, Gharbi K, Howard P, Ozaki A, et al. (2000) A microsatellite linkage map of rainbow trout (Oncorhynchus mykiss) characterized by large sex-specic differences in recombination rates. Genetics 155: 13311345. 52. Lercher MJ, Hurst LD (2003) Imprinted chromosomal regions of the human genome have unusually high recombination rates. Genetics 165: 16291632. 53. Sandovici I, Kassovska-Bratinova S, Vaughan JE, Stewart R, Leppert M, et al. (2006) Human imprinted chromosomal regions are historical hot spots of recombination. PLoS Genet 2: e101. doi.10.1371/journal.pgen.0020101 54. Henderson KA, Keeney S (2005) Synaptonemal complex formation: Where does it start? Bioessays 27: 995998. 55. Brown PW, Judis L, Chan ER, Schwartz S, Seftel A, et al. (2005) Meiotic synapsis proceeds from a limited number of subtelomeric sites in the human male. Am J Hum Genet 77: 556566. 56. Lynn A, Ashley T, Hassold T (2004) Variation in human meiotic recombination. Annu Rev Genomics Hum Genet 5: 317349. 57. Esch E (2005) Estimation of gametic frequencies from F2 populations using the EM algorithm and its application in the analysis of crossover interference in rice. Theor Appl Genet 111: 100109. 58. Colombo PC, Jones GH (1997) Chiasma interference is blind to centromeres. Heredity 79 (Pt 2): 214227. 59. Broman KW, Weber JL (2000) Characterization of human crossover interference. Am J Hum Genet 66: 19111926. 60. Hollingsworth NM, Brill SJ (2004) The Mus81 solution to resolution: Generating meiotic crossovers without Holliday junctions. Genes Dev 18: 117125. 61. Mercier R, Jolivet S, Vezon D, Huppe E, Chelysheva L, et al. (2005) Two meiotic crossover classes cohabit in Arabidopsis: One is dependent on MER3, whereas the other one is not. Curr Biol 15: 692701. 62. Lin S, Cheng R, Wright FA (2001) Genetic crossover interference in the human genome. Ann Hum Genet 65: 7993. 63. Broman KW, Rowe LB, Churchill GA, Paigen K (2002) Crossover interference in the mouse. Genetics 160: 11231131. 64. Tapper WJ, Ke X, Morton NE, Collins A (2002) Recombination, interference, and sequence: Comparison of Chromosomes 21 and 22. Ann Hum Genet 66: 7586. 65. Stahl FW, Foss HM, Young LS, Borts RH, Abdullah MF, et al. (2004) Does crossover interference count in Saccharomyces cerevisiae? Genetics 168: 3548. 66. Martini E, Diaz RL, Hunter N, Keeney S (2006) Crossover homeostasis in yeast meiosis. Cell 126: 285295. 67. King JS, Mortimer RK (1990) A polymerization model of chiasma interference and corresponding computer simulation. Genetics 126: 11271138.

PLoS Genetics | www.plosgenetics.org

1107

June 2007 | Volume 3 | Issue 6 | e106

76

2.3 Diversit des profils de crossover et de conversion gnique des points chauds miotiques chez

Arabidopsis thaliana.
2.3.1 Synthse de la publication
Ce travail porte sur ltude dtaille du point chaud de recombinaison miotique dnomm 130x , localis sur le bras court du chromosome 4 dA. thaliana. Ce point chaud a t originellement identifi par cartographie fine de CO dans une population F2 ColxLer (annexe). Aprs une phase prliminaire de mise au point de la technique de pollen typing et de tests des amorces utilises pour lamplification par PCR spcifique dallle de molcules uniques (section 4), 167 vnements de CO intervenant dans un intervalle de plus de 12 kb de la rgion 130x ont t caractriss chez le type sauvage. La frquence de CO atteint 0,3%. Le squenage de ces molcules a permis dtablir une carte de distribution des CO dans les deux orientations. Les points dchange se regroupent dans un intervalle de plus de 8 kb flanqu par des rgions dpourvues de CO. En outre la densit maximale de CO atteint 70 cM/Mb, soit 15 fois la moyenne du chromosome. Cette rgion correspond donc la dfinition usuelle dun point chaud de CO miotiques. Nanmoins, elle possde deux particularits qui la distinguent de la plupart des points chauds dcrits chez les mammifres : elle est trs large, et prsente un profil irrgulier de distribution des CO, avec des une alternance de pics et de valles . En outre, six insertions/dltions de taille importante (plus de 7 bp) jalonnent cette rgion. Les distributions de CO dans les orientations rciproques ( Col vers Ler =CL ou Ler vers Col =LC) paraissent lgrement dissemblables, mais aucune diffrence statistiquement significative na pu tre dtecte. Des CO ont t caractriss galement chez un mutant Atmsh4, qui est dfectif dans la formation des crossings-over de type I (interfrents). 56 molcules ont t obtenues, qui sont prsentes dans lADN gnomique de pollen une frquence de 1,53.10-4, soit 5,1 % du niveau du type sauvage. La distribution des points dchange est par ailleurs notablement diffrente de celle du type sauvage ; notamment 37 % dentre eux sont localiss droite de la rgion dfinissant le point chaud dans le type sauvage. De surcrot, des vnements de NCO ont t dtects par pollen typing la fois chez le type sauvage et chez le mutant Atmsh4, un site polymorphe localis au centre du point chaud 130x. 77

La conversion gnique survient une frquence de 1,9.10-4 chez le type sauvage (1/16 de la frquence de CO) et 3,4.10-5 chez Atmsh4 (22 % de la frquence de CO). La diminution de frquence des NCO dans Atmsh4 par rapport au type sauvage (5,6 fois) est donc trs notable, mais moindre que la diminution de frquence des CO (16 fois). Le petit nombre dvnements obtenu ne permet pas de dceler de diffrence entre les frquences de conversion dans les deux sens CL et LC. Les tracts de conversion les plus courts stendent sur moins de 870 pb, tandis que les plus longs dpassent 1400 pb.

2.3.2 Publication

78

Contrasted patterns of meiotic crossover and noncrossover events at Arabidopsis thaliana recombination hotspots

Jan Drouaud, Hossein Khademian, Laurne Giraut and Christine Mzard*

Institut Jean-Pierre Bourgin UMR1318 INRA-AgroParisTech Route de St-Cyr (RD10) 78026 Versailles Cedex France.
*

Corresponding author. E-mail: christine.mezard@versailles.inra.fr. Tel.: (33)130833739. Fax.: (33)130833319.

One-sentence summaries: MSH4 is involved in both gene conversion and crossovers at meiotic recombination hotspots in Arabidopsis

79

Abstract
The vast majority of meiotic recombination events cluster in narrow hotspots surrounded by large regions devoid of recombinational activity. Here, we show that such hotspots do also exist in plants. Characterization of three of them in Arabidopsis thaliana reveals several new features. First, crossovers (COs) concentrate in small regions of a few kilobases and their rates reach 9 to 18 times the genome average. Second, non-crossover (NCOs) events also occur in two hotspots but at very different levels (CO/NCO ratio of 1/1 and 16/1). Third, in the absence of the recombination protein MSH4, NCOs are considerably decreased (18 % of wild-type) and the CO distribution is also modified. We propose an unrevealed new role for MSH4 in the formation of NCOs that explains MSH4 localization in plants, fungi and mammals.

80

During meiosis, recombination is initiated by DNA double strand breaks (DSB) that are repaired, either as CO or as NCO, using the information of the homologous chromosome (1). CO distribution is not homogeneous along chromosomes at any scale. COs tend to be clustered in narrow regions of a few kilobases called hotspots (HS) where CO frequency is greatly enhanced compared to adjacent regions but rarely exceeds 1% of meioses in higher eukaryotes (2). In plants, meiotic recombination has been usually studied by segregation analysis of molecular, genetic or cytological markers in offspring (3). The possibility to characterize meiotic HS was severely limited by the number of individuals used with these approaches. Building a precise genetic map of A. thaliana chromosome 4 allowed us to identify a series of regions where COs are grouped in a few kilobases (4, 5). These were good HS candidates. Therefore, we set up a "pollen typing" molecular approach (fig. S1), based on the sperm typing technique developed previously in mammals (6), that allowed us to detect and characterize hundreds of CO molecules in two different HS candidate regions of chromosome 4: 14a on the long arm and 130x on the short arm. COs rate in pollen genomic DNA from a F1 Columbia x Landsberg erecta (Col x Ler) was 0.7 %, at 14a, and 0.3 % at 130x. All the COs molecules characterised (104 at 14a and 167 at 130x) contained a single transition between parental haplotypes. The 14a region exhibits two distinct peaks of COs (14a1 and 14a2), 1,535 bp and 3,649 bp wide, which are 3,083 bp apart with a in-between valley where just a few COs were detected; in contrast, CO frequency is null on each side of the regions (Fig. 1A). At 14a1 and 14a2, CO rates reach 316 and 154 cM/Mb, respectively (68 and 33 times the chromosome average). At the 130x, CO rate is also null on both sides of the region and reaches a maximum (close to the centre) at 70 cM/Mb, which is 15 times the chromosome average (Fig. 1B). Thus, both CO rates and CO distributions at those two sites indicate clearly the existence of HS in A. thaliana. However, HS can be different as indicated by the distribution of COs at 130x when compared to 14a1 and 14a2 HSs: it is quite irregular with alternating peaks and valleys,

81

and thus does not fit well a Gaussian curve, and COs spread over more than 8 kb. This unusually broad and irregular distribution, when compared to those reported for yeast and mammalian hotspots, could suggest that double-strand breaks (DSBs) are not clustered in a single narrow region, but are rather formed across a large area or in several discrete regions. On the other hand, the irregularities observed in the distribution could also be explained by the presence of several insertions/deletions - which are 13, 70, 10, 7 and 11 pb wide respectively - along the region (Fig. 1B). Such heterologies could either block branch migration of double Holliday junctions or channel recombination intermediates towards NCOs or exchanges between sister chromatids as suggested previously for the mice HS22 hotspot (7). The DSB-repair model for meiotic recombination and its derivatives predict that the broken chromatid is the recipient of the information coming from the homolog (1 , 8 , 9). If one homolog is preferentially involved in DSB formation, crossover asymmetry is observed, with the alleles of the less active homolog near the centre of HS being over transmitted (see (10) for a comprehensive discussion). At the 14a HS, some discrepancies between CO distributions in reciprocal orientations Col to Ler or Ler to Col (i.e. CtoL and LtoC) can be observed (Fig. 2A). The difference is highly significant at 14a1 (p-val 0.00111) while it is only weakly significant (p-val 0.0734) for the 14a2 hotspot, probably due to the lower CO number. In both cases, LtoC CO breakpoints are shifted to the left when compared to CtoL CO breakpoints (Fig. 2A). These patterns are consistent with the hypothesis that the Ler allele has a stronger initiation activity than the Col allele at these HS. In contrast, at the 130x HS both alleles are equally proficient for initiating recombination (Fig. 2B). Interestingly, the centers of all three hotspots lie close to gene promoters, which are active in A. thaliana meiocytes (3 to 4 times above the average transcription level, C. Horlow, personal communication). A similar localization has been noticed for most DSB hotspots in Saccharomyces cerevisiae (11, 12). In human, genome wide characterization of historical CO

82

hotspots, through the analysis of linkage disequilibrium, provides evidence that recombination mainly occurs near genes but outside the transcribed region (13). Nevertheless, in mammals as well as in yeasts, no clear relation between recombination and transcription has been related until now. Meiotic DSBs are repaired either as CO or as NCOs. However, their relative frequencies vary from one HS to another (2, 14, 15). Until now, the underlying mechanical basis of these fluctuations remains totally unknown. In plants, very few meiotic NCOs have been characterized (5, 16) because of the difficulty of detection when they are not linked to a phenotypic change. We characterized NCO events at both 14a1 and 130x HS, with different molecular approaches. At the 130x, using a polymorphic site (#44) located at the center of the CO distribution three LtoCtoL and two CtoLtoC events were found among 26,840 molecules (Fig. 3B) (see Material and Methods). The observed NCO frequency at this marker is about 1/5,368, which is 16 times less than the overall CO frequency. Four NCO tracts out of five extend on the right to the neighbouring polymorphism (89 bp away), while polymorphisms on the left are co-converted in two tracts only, but over a greater distance (553 bp and 1311 bp, respectively). Considering the unusually broad distribution of COs, DSBs then NCO events could occur all over the large central region of the 130x HS. Therefore, the global rate might be actually higher than 1/5,368. At 14a1, among 3,052 molecules tested, two LtoCtoL and six CtoLtoC NCO events were detected at a SNP (#35) very close to the center of the HS (Fig. 3A). Thus, the NCO frequency (1/382) is about half of the overall CO frequency estimated with the pollen typing approach (1/150). Given the SNP location, close to the center of this narrow (1.5 kb) HS, it is likely that #35 is the most frequently converted SNP in 14a1. All NCO events were restricted to the SNP #35, i.e. without co-conversion of left and right flanking markers, which are located 111 and 482 bp apart respectively. Therefore, the mean NCO tract length is less than 593 bp. Importantly, our

83

results show that meiotic NCOs can occur at high rate in plant HSs and their mean lengths are short (a few hundreds of bp or even less). In A. thaliana, on the basis of chiasma counts, it is assumed that 85% of COs belong to the interference dependent pathway (class I), while the remaining 15% are interference-free (class II)(17). To test the contribution of both CO pathways, we analyzed their distribution at the 130x HS in an Atmsh4 mutant background in which interfering COs are absent. As seen in Figure 4, the distribution of CO exchange points along the 130x HS does depart notably from that seen in the wild-type: 37 % of COs in msh4 are located on the right part of the region beyond the most external CO described in wild-type. Interestingly, we observed one long chimeric CO with multiple exchange between parental molecules while it was never detected in wild type. Moreover, we detected a dramatic decrease of CO frequency in Atmsh4 (5.2 % of wild-type level). Hence, CO rate reduction in Atmsh4 is about 3 times greater at the 130x HS than for the whole genome. This is a strong indication that the proportion of MSH4 dependent COs can vary markedly in A. thaliana, both along chromosomes and at a hotspot scale. Next, we tested the NCO frequency in the Atmsh4 mutant background. Among 118,762 molecules, one LtoCtoL and three CtoLtoC events were detected (Fig. 4). NCO rate at this site is 1/29,690, which is 18 % of wild-type. Therefore, NCO frequency at #44 is considerably decreased in Atmsh4. Since the number of RAD51 foci is similar in wild-type and Atmsh4 (17), we presume that DSBs formation is unaltered by this mutation. Hence, our results show clearly that recombination intermediates that are committed to form class I COs in wild-type, are not channelled to NCOs in Atmsh4. The fate of these intermediates is unknown, but they are possibly repaired using the sister chromatid. Interestingly, NCO frequency is found decreased by the msh4 mutation in A. thaliana, while it is claimed unaffected in S. cerevisiae (15).

84

In both A. thaliana and mice, the MSH4 protein locates in numerous foci along chromosome axes as early as mid leptotene and then the number of foci decreases to zero at the end of pachytene (17, 18). In both cases though, the maximum number of MSH4 foci far exceeds the number of COs. A recent study in Sordaria macrospora shows that MSH4 is present at virtually all sites of interaction between homologs (COs and NCOs) at the onset of zygotene, and that it plays a role in the orderly progression of the pairing and synapsis processes (19). Additional insights emerge from our discovery of both the variation of COs from one HS to another and the decrease in both COs and NCOs in absence of MSH4. Moreover, it provides a molecular explanation to these cytological observations. We propose that in higher eukaryotes, beyond its role in class I CO formation, MSH4 plays a role in the formation or in the stabilization of at least some of recombination intermediates leading to NCOs.

85

References and Notes


1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. J. W. Szostak, T. L. Orr-Weaver, R. J. Rothstein, F. W. Stahl, Cell 33, 25 (1983). N. Arnheim, P. Calabrese, I. Tiemann-Boege, Annu Rev Genet 41, 369 (2007). A. D. Muyt, R. Mercier, C. Mezard, M. Grelon, Genome Dyn 5, 14 (2009). J. Drouaud et al., Genome Res 16, 106 (Jan, 2006). C. Mezard, Biochem Soc Trans 34, 531 (Aug, 2006). A. J. Jeffreys, J. Murray, R. Neumann, Mol Cell 2, 267 (1998). P. R. Bois, Mol Cell Biol 27, 7053 (Oct, 2007). N. Hunter, N. Kleckner, Cell 106, 59 (2001). T. Allers, M. Lichten, Mol Cell 8, 225 (2001). F. Baudat, B. de Massy, Chromosome Res 15, 565 (2007). J. L. Gerton et al., Proc Nat Acad Sci Usa 97, 11383 (2000). H. G. Blitzblau, G. W. Bell, J. Rodriguez, S. P. Bell, A. Hochwagen, Curr Biol 17, 2003 (Dec 4, 2007). S. Myers, L. Bottolo, C. Freeman, G. McVean, P. Donnelly, Science 310, 321 (Oct 14, 2005). F. Baudat, B. de Massy, Chromosome Research 15, 565 (2007). E. Mancera, R. Bourgon, A. Brozzi, W. Huber, L. M. Steinmetz, Nature 454, 479 (Jul 24, 2008). K. E. Francis et al., Proceedings of the National Academy of Sciences of the United States of America 104, 3913 (Mar, 2007). J. D. Higgins, S. J. Armstrong, F. C. Franklin, G. H. Jones, Genes Dev 18, 2557 (Oct 15, 2004). B. Kneitz et al., Genes Dev 14, 1085 (May 1, 2000). A. Storlazzi et al., Cell 141, 94 (2010). We thank all members of our laboratories for discussions, Mathilde Grelon, Eric Jenczewski, Raphal Mercier, Fabien Nogu, Olivier Loudet and Denise Zickler for helpful comments on the manuscript, Aurlie Brard and Dominique Brunel for plasmid genotyping, Christine Horlow for providing data on gene expression in meiocytes before publication. This study was supported by grants from Europe (FP7 MEIosys) and Agence Nationale de la Recherche (grant ANR GNP05040G).

86

Figure legends
Figure 1. Distribution of CO rates across the 14a (A) and 130x (B) hotspots. 104 COs were analyzed at 14a (62 at 14a1 and 42 at 14a2) and 167 at 130x Grey dotted vertical lines: median positions of the hotspots. Black diamonds: positions of insertions-deletions Gene structures along the hotspot regions are displayed on top of graphical areas (light hashed grey: exons, dark grey: UTR, broken lines: introns).

Figure 2. Distributed and cumulated CO rates for reciprocal orientations at 14a (A) and 130x (B). 29 CtoL and 33 LtoC COs were analysed at 14a1, 23 and 19 at 14a2, and 75 and 92 at 130x. For all three hotspots and both reciprocal orientations, cumulated (from left to right) relative CO rates were scored at successive polymorphic sites along the region. Blue curves: cumulated relative CtoL CO rates. Red curves: cumulated relative LtoC CO rates. Blue histogram: distribution of CtoL CO rates. Red histogram: distribution of LtoC CO rates. Grey dotted vertical lines: median positions of the hotspots.

Figure 3. NCO tracts in wild-type at 14a1 (A) and 130x (B). (A) NCO tracts at 14a1. The cloned DNA region used in genotyping is enlarged (see Material and Methods). (B) NCO tracts at 130x. The central part of the 130x HS is enlarged.

The two SNPs (#35 for 14a1 and #44 for 130x) used for isolating NCOs are indicated by vertical arrows. The polymorphisms in the two regions are indicated by blue vertical dashes. Thin horizontal lines: parental DNA, black Col, grey Ler. Thick horizontal lines: NCO tracts, black Col, grey Ler. Grey dotted vertical lines: median positions of the hotspots. 87

Figure 4. Distribution of CO rates and NCO tracts in Atmsh4 at 130x. 56 COs were analyzed at 130x. (A) (B) Distribution of CO rates. Grey area: Atmsh4. Dashed line: wild-type. NCO tracts. Black vertical lines: polymorphisms. The SNPs #44 used for isolating NCOs is indicated by a vertical arrow. The polymorphisms are indicated by blue vertical dashes. Thin horizontal lines: parental DNA, black Col, grey Ler. Thick horizontal lines: NCO tracts, black Col, grey Ler. For two NCOs the exact position of the left end of the converted tract is not known. Grey dotted vertical lines: median positions of the hotspots.

88

350 300 CO rate (cM/Mb) 250 200 150 100 50 0


2 69 6 1 3 69 6 1

AT4G35070

AT4G35080

14a1

14a2

4 69 6 1

5 69 6 1

6 69 6 1

7 69 6 1

8 69 6 1

9 69 6 1

0 70 6 1

chromosome coordinate (kb)

100 80 CO rate (cM/Mb) 60 40 20 0 172

AT4G00400

AT4G00416

AT4G00420

130x

173

174

175

176

177

178

179

180

181

182

183

chromosome coordinate (kb)

Figure 1. Distribution of CO rates across the 14a and 130x hotspots

89

300 200 100 0

14a1

14a2

75 50 25

cumulative CO rate (%) cumultaive CO rate (%)

400 CO rate (cM/Mb)

LtoC CtoL

100

0 92 693 694 695 696 697 698 699 700 6 16 16 16 16 16 16 16 16 16 chromosome coordinate (kb)

100 80 60 40 20

100

130x
LtoC CtoL

CO rate cM/Mb

80 60 40 20

0 0 173 174 175 176 177 178 179 180 181 182 chromosome coordinate (kb)

Figure 2. Cumulative CO rates for reciprocal orientations


90

14a

16693

16694

16695

kb

130x

B
177 178 179 kb

Figure 3. Gene conversion tracts in wild-type.

91

chromosome coordinate (kb) 173 174 175 176 177 178 179 180 181 182 183 184 185 186 4 80

crossover rate (cM/Mb) Atmsh4

40

20

177

178

179

kb

Figure 4. Distribution of CO rates and gene conversion tracts in Atmsh4.

92

crossover rate (cM/Mb) wild-type

60

Supplemental Materials

Material and Methods


Plant material The Arabidopsis thaliana accessions Columbia-0 (186AV) and Landsberg erecta (213AV) were obtained from the Centre de Ressources Biologiques at the Institut Jean Pierre Bourgin, Versailles, France. All plants were grown in the greenhouse under standard conditions. Pollen harvesting Whole inflorescences from hybrid plants were harvested in 10% saccharose, and crushed in a Waring Blendor (two 4 sec pulses at full speed). The homogenate, containing intact microspores and pollen grains, was next filtered through a 80 m steel mesh. Pollen was pelleted at 350 g, washed with 10% saccharose, pelleted at 100 g and stored at -20C until DNA extraction. Extraction of pollen genomic DNA Pollen grains and microspores were resuspended in four volumes of (50mM Tris-Cl pH8, 100 mM NaCl, 1 mM EDTA, 1 % SDS, 1 mM DTT, 20 g/ml proteinase K) and incubated at 65C for three hours with gentle shaking every 15 minutes. Then, ten 1 mm diameter glass beads were added into the suspension, and pollen grains were disrupted by mixing with a vortex at full speed for 1 to 3 minutes (cell disruption was monitored with a microscope every 30 sec). One volume liquid phenol (buffered with 1M Tris-Cl pH8) was added and tubes were rocked for 30 min at 4C. After centrifugation, the supernatant was recovered and nucleic acids were precipitated by sodium acetate and ethanol addition. Genomic DNA was next dissolved in (10 mM Tris-Cl pH8, 1 mM EDTA, 100 g/ml RNAseA) and incubated at room temperature for 15 min. Four volumes of (5M guanidine isothiocyanate, 50 mM Tris-Cl pH 8) were then added, and DNA was purified with DNeasy minicolumns (Qiagen ref. 69106). Extraction of leaf genomicDNA 93

Young leaves were frozen with liquid nitrogen in a mortar. An equal amount of poly(vinylpolypyrrolidone) (Sigma-Aldrich ref. 81385) was added and the material was ground to a fine powder with a pestle. The powder was transferred to a 30 ml centrifuge tube and thoroughly resuspended in hot (65C) lysis buffer (2% cetyltrimethylammonium, 100 mM Tris-Cl pH 8, 1.4 M NaCl, 20 mM EDTA, 10 mM 2-mercaptoethanol) (5ml for each gram of fresh tissue). The homogenate was incubated at 65C for 30 min, with occasional gentle shaking. After cooling to room temperature, the lysate was extracted by an equal volume of (chloroform/isoamylalcohol 24v/1v), then centrifuged at 15000 g for min. The upper aqueous phase was recovered and 0.7 volume isopropanol was added. Nucleic acids were pelleted by centrifuging at 15000 g for 10 min, then washed with 2 ml of 70 % ethanol, then pelleted again by centrifuging at 15000 g for 2 min. The pellet was dried at room temperature, then dissolved in 40 l of (10 mM Tris-Cl pH 8, 1 mM EDTA) per gram of fresh material. Four volumes of (5M guanidine isothiocyanate, 50 mM Tris-Cl pH 8) were then added, and DNA was purified with DNeasy minicolumns (Qiagen ref. 69106). Quantification of genomic DNA PCR reactions were performed in 20 l of the following buffer: 45 mM Tris-Cl pH 8.8, 11 mM (NH4)2SO4, 4.5 MgCl2, 6.7 mM 2-mercaptoethanol, 4.4 m EDTA, 113 g/ml BSA, 0.8 mM each dNTP, 0.4 M each oligonucleotide, 1 U Taq DNA polymerase, 0.1 U Pfu DNA polymerase. Whenever less than 100 pg/l of genomic DNA was used, carrier DNA (from herring sperm) was added into the reactions (1 ng/l). Allele-specific oligonucleotides (ASOs) were designed so that their Tm differential (between matching and mismatching templates) was more than 8C, and their optimal annealing temperature was around 60C. The quantification of amplifiable molecules in genomic DNA extracts was performed on series of dilutions through two rounds of PCR, using nested ASOs. The product of the first PCR was diluted 1/1000 in the second reaction. The thermal cycling profile of the reactions is: ((92C;2

94

min){(92C;20 sec)(Tm;30 sec)(68C;30 sec + 45 sec/kb)}x30(68C;90 sec/kb)(4C;)), where Th is the annealing temperature. For calculating the concentration of amplifiable molecules used in the first PCR, it was assumed that the initial number of target molecules in a reaction follows a Poisson distribution. Hence, after the second PCR, the proportion of negative wells among a set of aliquot reactions is approximated by e-m, where m is the mean initial number of DNA molecules per well. Parental molecules were thus quantified using ASOs all specific to either Col or Ler DNA. CO molecules were amplified with ASOs specific to either Col or Ler on one side, and Ler or Col respectively on the other side (fig. S1). Oligonucleotides data (sequence and genomic coordinates) are listed in table S1. Pairs of oligonucleotides used for allele-specific PCR are listed in table S2. Control reactions assessing the specificity of CO amplification. The meiotic origin of CO molecules which were amplified by allele-specific PCR was assessed with control experiments, carried out in parallel with pollen and leaf (somatic) DNA. Figure S2 features the result of such control reactions performed with ASO pairs 130x0LeL1/130x52CoR1 then 130x7LeL5/130x47CoR2 (table S2). It can be noticed that, using the same amounts of DNA, CO molecules could be detected in pollen, but none in leaf. Mapping of CO exchange points For mapping CO exchange points, series of aliquot reactions were carried out, which were predicted to contain an average of less than 0.2 CO molecule, so that more than 90 %
0.2 e 0.2 1 0.2 e 0.2 of positive reactions issued from a single CO molecule. PCR products were

then sequenced in order to locate exchange points from single CO molecules. Plotting CO frequencies across hotspots CO rate FCO across a hotspot is calculated as the average between CO rates in both orientations. Given NCL the total number of mapped CO breakpoints in the Col to Ler orientation, n1CL, n2CL, .. , nkCL the number of CO breakpoints mapped in intervals 1,2,..,k respectively, l1CL, l2CL, .. , lkCL 95

the size (in pb) of intervals 1,2,..,k respectively, the Col to Ler CO rates (in cM/Mb) in intervals

n kCL n n 1CL FCO 10 2 2CL FCO 10 2 FCO 10 2 N N N 1, 2, .. , k, are CL -6 respectively. The same , , CL -6 , CL -6 10 l k 10 l 2 10 l 1 rationale is used for plotting the distribution of CO in the Ler to Col orientation. At the 130x HS, the analysis was performed on two overlapping intervals, thus providing two distributions of CO breakpoints with one common interval. As expected, the CO frequencies in this common interval are similar for the two distributions (two-tailed Fishers exact test p-value = 0.23 for wild-type), then the average values were considered for merging the two overlapping distributions.
Statistical testing for differences between distributions of CO in reciprocal orientations.

At a given hotspot CO breakpoints located on each side of the median position were grouped separately for the two reciprocal orientations (Col to Ler CL and Ler to Col LC), thus providing four numbers COCLleft, COCLright, COLCleft, COLCright. These numbers were grouped in a contingency table for testing the association between left/right and CL/LC classifications using the two-tailed Fishers exact test.
Characterization of NCO events at 130x hotspot

NCO molecules at polymorphism #44 in 130x hotspot were detected using a strategy adapted from (S1). The outline of this approach is described in figure S3. Oligonucleotides used for these experiments are listed in table S3.
Characterization of NCO events at 14a hotspot

A Col or Ler-specific 7.3 kb genomic DNA fragment was amplified by two rounds of PCR with oligonucleotides pairs 14a8-14a63 and 14a9-14a54 (table S2), as described previously. The PCR products were then used to amplify a 2.5 kb fragment encompassing the 14aHS region, with oligonucleotides 14aKH1L and 14aKH2bisR (table S1). The PCR product were then digested with restriction enzymes BglII and XbaI, and the resulting 2.1 kb product was ligated into pCRIITOPOblunt between the BamHI and XbaI unique sites, using standard procedures. The ligation product was then used to transform DH10B E. coli strain by electroporation.
96

Transformed cells were spread onto LB agar plate containing 100 g/ml carbenicillin, 0.2 mM IPTG and 40 g/m X-gal. Following blue/white screening, individual colonies were transferred into 200 l of LB medium containing 100 g/ml carbenicillin in 1 ml MASTERBLOCK microplates (Greiner Bio-One ref. 780215) and grown with gentle shaking at 37C for 16 hours. Then, 100 l of cell cultures were transferred in 96 well V-bottom microplates, and spun down at 3200 g for 10 min. Cell pellets were resuspended in 100 l of sterile water. For each group of four 96 wells plates, eight wells were used for control reactions: four contained sterile water, four contained E.coli cells transformed with a plasmid containing parental DNA (two wells for Col and two wells for Ler). Bacterial clones were then genotyped using the Chemicon Amplifluor SNPs Genotyping System. For this purpose, oligonucleotides either specific to each parent at polymorphism #35, or non-specific, were designed using the Amplifluor AssayArchitect software (https://apps.serologicals.com/AAA/): 1201-Y-Col GAAGGTCGGAGTCAACGGATTTTTCTTTTCGGCTTTTTCGG, 1201-Y-Ler GAAGGTGACCAAGTTCATGCTATTTCTTTTCGGCTTTTTCGA and Y-1021-Y-Rev GTGCCCTGCCATAATTCCACT. Genotyping was then performed as described in (S2).

Supplemental references
S1. S2. F. Baudat, B. de Massy, in Methods Mol Biol. (2009), vol. 557, pp. 305-22. S. H. Ng, E. Parvanov, P. M. Petkov, K. Paigen, Genomics 92, 204 (Oct, 2008).

Supplemental figure legends


Figure S1. Outline of the pollen typing method.

(A) Within pools of hundreds to thousands molecules of genomic DNA extracted from Col x Ler hybrid pollen, single CO molecules are amplified by two rounds of PCR using oligonucleotides (triangles) specific to Col or Ler (white and black respectively) polymorphisms (circles).
97

(B) CO molecules are sequenced, in order to locate exchange points (thick grey lines), allowing to measure and plot CO rates across the hotspot.

Figure S2. Detection of CO molecules in genomic DNA from hybrid leaf or pollen.

PCR is performed with allele-specific oligonucleotides designed for the amplification of CO molecules (see Material and Methods), using decreasing amounts of genomic DNA (the number of template molecules is indicated on the photograph) extracted from F1 Col x Ler hybrid plants, either from leaf (two top rows) or pollen (two bottom rows). Eight aliquot reactions are carried out for each dilution. Faint bands can be detected for high amounts of leaf DNA (16384 template molecules and more), which are all non-specific (data not shown) PCR products. No signal can be seen for lower DNA amounts. Conversely, strong and specific amplification of CO molecules is obtained for lower pollen DNA amounts (4096 molecules and less).

Figure S3. Outline of the PCR-based strategy for NCO characterization

A series of aliquot PCR is performed with an allele-specific oligonucleotide (P1) on one side of the studied region and a non-specific oligonucleotide (P2)on the other side, using an amount of template genomic DNA predicted to contain less than one recombinant (either CO or NCO) molecule per reaction on the average. This allows amplifying the molecule types displayed in the lower part of the figure, named parental, CO left, CO right and NCO. The product of the first PCR is used as a template for carrying out four separate nested PCR, using ASOs P3, P4, P5, P6, P7 and P8. Depending both on the type of target molecule present in the first PCR product and the combination of ASOs, the second PCR can yield a product or not. Hence, the amplification pattern of the four nested PCR is used as diagnostic data for the identification of the type of recombinant molecule initially present in the reaction.

98

Supplemental tables

Oligonucleotide name 130x0CoL1 130x0LeL1 130x7CoL4 130x7LeL5 130x47CoR2 130x47LeR4 130x52CoR1 130x52LeR2 130x43CoL1 130x42LeL1 130x44CoL4 130x44LeL4 130x72CoR2 130x72LeR2 130x78CoR2 130x78LeR3 14a8CoL3 14a5LeL3 14a9Col2 14a9LeL2 14a23CoL1 14a23LeL1 14a54CoR2 14a54LeR2 14a63CoR3 14a63LeR3 14aKH1L 14aKH2bisR

Genomic coordinates CGCACACTCTTATCCAACA 171696..171704 CGCACATTCTTTTCCAACT 171696..171704 GCCTAATGTTCCCATGGATCTC 173199..173220 TGCCTAATGTTCCCATATCGAAG 173198..173220 TGAGTCAATTAACGTGGTTTGCA 179051..179032 GTCGATGTCGTTGGTTTTCG 179049..179032 TGTCAGAGATATTCTAGACATGTGA 180432..180408 TGGGATGAATTTTGTCAGAGCTATGA 180444..180408 GAAGAATGGATTATATAATACAGTACAGTATAT 178138..178170 GGTTACAGAGAATTGTAGAATGTTAC 178103..178128 AGTGGTCACTGTCGCTATAG 178344..178363 AAGTGGTCACTGTCGTTATAC 178343..178363 GGGCCAATACATTGGAAACA 185687..185668 CAAGGGCCAAGACCTCAATC 185690..185666 CTTTTGTCTATGTAAAAACTATGTG 186013..185989 GGTTGGCCTCTTCTG 186004..185991 TATGCAACAGCAACACG 16691260..16691275 CCCCCCGCATGCTTTTACATTATA 16691245..16691263 ACCGTTGCACTTTCCTT 16692501..16692519 CCGAAACCGTTGCACTATG 16692500..16692530 CCCCCGATTTATTCATACATATC 16692993..16693014 CCCCCGATTTATTCGTACATATT 16692993..16693014 GGAGAGCTAATGCAGGC 16699799..16699783 GCAGAGCCAATGCGGGT 16699799..16699783 CCTTTTGTTCGTACAGGTG 16700136..16700120 CCCCCATTTTGTTCGCAAAGGTA 16700137..16700120 ATCTTATAAACGTTATTGTCA 16692865..16682885 TTCGCCCGGCAAACTATTCC 16695220..16695202 Sequence 5'->3'

Table S1. Oligonucleotides data.

99

Oligonucleotide pair for first PCR 130x0CoL1/130x52CoR1 130x0LeL1/130x52LeR2 130x0CoL1/130x52LeR2 130x0LeL1/130x52CoR1 130x43CoL1/130x78CoR2 130x42LeL1/130x78LeR3 130x43CoL1/130x78LeR3 130x42LeL1/130x78CoR2 14a9Col2/14a63CoR3 14a9LeL2/14a63LeR3 14a8CoL3/14a63LeR3 14a5LeL3/14a63CoR3

Annealing temperature (C) 59 59 59 59 58 58 58 58 58 58 61 61

Oligonucleotide pair for second PCR 130x7CoL4/130x47CoR2 130x7LeL5/130x47LeR4 130x7CoL4/130x47LeR4 130x7LeL5/130x47CoR2 130x44CoL4/130x72CoR2 130x44LeL4/130x72LeR2 130x44CoL4/130x72LeR2 130x44LeL4/130x72CoR2 14a23CoL1/14a54CoR2 14a23LeL1/14a54LeR2 14a9Col2/14a54LeR2 14a9LeL2/14a54CoR2

Annealing Target temperature molecule type (C) 59 parental 59 parental 59 CO 59 CO 58 parental 58 parental 58 CO 58 CO 58 parental 58 parental 58 CO 58 CO

Table S2. Oligonucleotides used for allele-specific PCR.

100

First oligonucleotide Nickname P1 P3 P3 P5 P7 Name 130x0CoL1 130x7CoL4 130x7CoL4 130x44LeL4 130x44CoL4

Col to Ler Second oligonucleotide Nickname P2 P4 P8 P6 P8 Name 130xUR13 130x44LeR4 130x52LeR2 130x52CoR1 130x52LeR2

Annealing temperature (C) 59 63 60 63 60

First oligonucleotide Nickname P1 P3 P3 P5 P7 Name 130x0LeL1 130x7LeL5 130x7LeL5 130x44CoL4 130x44LeL4

Ler to Col Second oligonucleotide Nickname P2 P4 P8 P6 P8 Name 130xUR13 130x44CoR1 130x52CoR1 130x52LeR2 130x52CoR1

Annealing temperature (C) 60 62 64 61 61

Table S3. Oligonucleotides used for NCO characterization at 130x.

101

Col X Ler A

B cM/Mb

Kb

Figure S1. Outline of the pollen typing method.

102

32768

16384

8192 Leaf

4096

2048

1024

4096

2048

1024 Pollen

512

256

128

Figure S2. Detection of CO molecules in genomic DNA from hybrid leaf or pollen.

103

P1

P2

after PCR with primers P1 and P2 pair of ASOs


P3 P7 P5

P3-P8 parental type of molecule CO right CO left NCO


P4 P6 P8

P3-P4

P5-P6

P7-P8

+ + -

+ +

+ -

Figure S3. Outline of the PCR-based strategy for NCO characterization.

104

3 Discussion et perspectives
3.1 Htrochiasmie et variations de longueur du complexe synaptonmal
La signification volutive de lhtrochiasmie a t abondamment discute dans de nombreux articles (cits notamment dans (Burt, Bell et al. 1991; Lenormand 2003; Lenormand et Dutheil 2005)). Ce sujet sort du cadre de la thse, et ne fait dailleurs lobjet daucun consensus : ce qui serait vrai dans les mammifres euthriens ne lest pas chez les marsupiaux, et des plantes assez troitement apparentes comme A. thaliana, la moutarde, le chou et le colza prsentent des variations discordantes du ratio des nombres de crossings-over dans les deux sexes. Au regard de ce qui a t prsent prcdemment (section 1.3.1.1. de lintroduction, et section 2.2.3. des rsultats), concernant la covariation entre la taille du complexe synaptonmal et le nombre de crossings-over, il apparat que la compaction des chromosomes diffre dans les mioses des deux sexes dans de nombreuses espces dont A. thaliana, et que cest possiblement cela qui dtermine la variation du nombre de crossings-over. Une synthse des donnes bibliographiques sur cette question de la longueur du complexe synaptonmal a t ralise en 2006 (Kleckner 2006). Elle rvle que lespacement des boucles de chromatine le long du complexe synaptonmal est assez bien conserv parmi les eucaryotes. Pour une taille de gnome / quantit dADN donne, il existe donc une corrlation inverse entre la taille des boucles, et la longueur du complexe synaptonmal. Sur un plan mcanique, on ignore quelle variable peut dterminer lautre, et de quelle manire. En particulier, aucune squence dADN consensus susceptible de permettre lancrage des boucles de chromatine na jusqu prsent pu tre clairement mise en vidence. Nanmoins on peut observer une corrlation entre la taille du complexe synaptonmal et le volume nuclaire, faible entre diffrentes cellules chez Schizophyllum commune (Carmi, Holm et al. 1978), ou C. elegans (Goldstein 1986), plus forte entre diffrentes souches chez S. cerevisiae (Moens et Ashton 1985). Il semblerait donc que la taille des boucles soit une variable dajustement , et que ce soit plutt la longueur des axes ou des bivalents qui importe, a priori en fonction du volume du noyau. On peut imaginer que des contraintes striques sexercent sur la dynamique des chromosomes, qui est susceptible dtre mise en jeu diffrentes tapes cls du processus de recombinaison miotique soit au stade zygotne, lors de la mise en place du synapsis (qui est vraisemblablement coupl linitiation de la recombinaison, voir section 1.2.1.1.3.2.), soit au stade pachytne, lors de

105

la maturation des intermdiaires de recombinaison (Dawe, Sedat et al. 1994; Scherthan, Wang et al. 2007).

3.2 Interfrence et distribution des crossovers le long du chromosome 4


La diffrence trs nette observe entre les distributions de CO le long du chromosome 4 dans les mioses des deux sexes peut sexpliquer simplement par laction de linterfrence. En effet, la proportion thorique de bivalents prsentant des crossings-over multiples (section 2.2.2.) est faible dans la miose femelle (7,3 %), et leve dans la miose mle (64 %). En outre, la distribution thorique du nombre de crossings-over par bivalent dans la miose femelle est compatible avec la prsence presque systmatique dun crossing-over obligatoire exerant une interfrence totale, auquel sajoute une faible proportion de crossings-over non interfrents, intervenant alatoirement. Dans la miose mle par contre, linterfrence nest pas totale, de sorte que les crossings-over multiples, qui sont supposs dpendre de la voie I pour la plupart, sont nombreux et ont tendance se repousser mutuellement. On peut attendre en consquence une augmentation de la frquence des CO issus de crossings-over multiples chaque extrmit du chromosome, tandis que les CO issus de crossings-over simples se distribuent vraisemblablement de la mme manire dans les deux sexes sur une large portion centrale du chromosome ((Drouaud, Mercier et al. 2007) figure 1C). Dans cette hypothse, les CDB initiatrices seraient distribues plus uniformment le long du chromosome que ne le sont les CO, et, corrolairement, la balance CO/NCO serait suprieure aux extrmits du chromosome, comparativement la partie mdiane, dans la mesure o les vnements de recombinaison miotique impliquent principalement des chromatides homologues. Ltude de la distribution lchelle du gnome entier dA. thaliana des vnements de CO ET de NCO dans une srie de produits de miose (ttrades) permettrait de tester cette hypothse.

3.3 Variations de la concidence des crossovers et intensit de linterfrence le long du chromosome 4


La variation conjointe de la concidence des CO et de la taille physique des intervalles considrs pour la calculer peut signifier : soit que la force de linterfrence est correctement mesure par la concidence, et quelle varie en fonction de la distance physique, mais pas de la distance gntique, sparant les crossingsover. 106

soit que la variation de concidence est dtermine par une variation spatiale (i.e. le long du chromosome) de la proportion des crossings-over de type II, auquel cas linterfrence ne peut tre mesure fidlement par une analyse de concidence. Cette dernire ventualit, suggre par F. Stahl (communication personnelle) semble impossible exclure si la proportion globale des crossings-over de type II est leve (de lordre de 15% (Higgins, Armstrong et al. 2004)). On peut en effet imaginer que la portion centrale du bras long, qui prsente une concidence leve de CO ((Drouaud, Mercier et al. 2007) figure 6A), est pauvre en crossings-over de type I du fait de linterfrence qui les repousse vers les extrmits, et donc relativement plus riche en crossings-over de type I insensibles linterfrence. En revanche, cette hypothse parat dautant moins vraisemblable que la proportion globale des crossings-over de type II est plus faible. Or, cela pourrait tre le cas. Dans un hybride F1 ColxLer, le nombre total de crossings-over par miose (dduit du nombre de CO mesurs dans une population de grande taille, par une approche similaire celle prsente dans la section 2.2.3.), atteint 11,2 (L. Giraut, communication personnelle). Par ailleurs, le nombre de foyers MLH1 mesur au stade diplotne dans un hybride ColxLer avoisine 10,3 (L. Chelysheva, communication personnelle). Il semble probable que ces dcomptes, raliss dans un type sauvage par des mthodes peu susceptibles dintroduire un biais, refltent plus correctement la ralit biologique. La proportion des crossings-over de type II serait alors proche de 7 %. En outre, dans la miose femelle les chiasmas en excs du crossing-over obligatoire reprsenteraient 8,2 % du total (section 2.2.2.). Il semble raisonnable de supposer que la proportion des crossings-over de type II est au plus gale cette valeur, peut tre dans la miose mle aussi bien. En ltat actuel des connaissances, il semble difficile de trancher cette question. Deux stratgies peuvent tre envisages qui consistent dterminer la distribution fine des CO de type I et de type II, respectivement. Il semble que la mutation du gne AtMUS81 ne suffise pas abolir tous les CO de type II, puisque des crossings-over apparemment non-interfrents sont encore observs dans le double mutant Atmsh4/AtMUS81 (Higgins, Buckling et al. 2008). Il parat donc plus judicieux de raliser une tude de la distribution fine des CO (de type II) dans le mutant Atmsh4, mais la faiblesse du taux rsiduel de CO (voir plus haut) constitue un inconvnient majeur puisquen contrepartie la taille des populations analyser doit tre beaucoup augmente, afin de maintenir la puissance de lanalyse statistique un niveau convenable.

107

3.4 Distribution atypique des crossovers au point chaud 130x


La largeur et lirrgularit de la distribution des CO dans la rgion 130x peuvent tre expliques par deux raisons qui ne sont pas mutuellement exclusives : Il existe non pas une mais plusieurs zones discrtes dinitiation de la recombinaison miotique, distribues sur plusieurs kilobases, et qui sont trop proches les unes des autres pour que les CO quelles engendrent se distribuent dans des pics distincts. Les polymorphismes de type insertion/dltion (INDEL) bloquent la migration de branche des intermdiaires de recombinaison, et/ou favorisent lvolution de ces derniers en vnements de NCO ou dchange avec une chromatide sur. Afin dapporter des arguments exprimentaux en faveur de lune ou lautre hypothse, il conviendrait dtudier le point chaud 130x dans dautres contextes hybrides, (i) qui ne prsenteraient pas les INDEL observs dans ColxLer, et (ii) qui seraient par ailleurs susceptibles de ne pas possder le mme niveau dactivit toutes les zones ventuelles de formation des CDB initiatrices. Cette situation serait alors similaire celle des points chauds S1 et S2 humains qui sont localiss proximit lun de lautre (2 kb) mais sont nanmoins rguls distinctement dans des individus diffrents (Jeffreys et Neumann 2009). Des groupes de points chauds physiquement trs proches les uns des autres ont t dcrits chez lhumain : DNA1-DNA2-DNA3, DMB1-DMB2 (Jeffreys, Kauppi et al. 2001), NID2a-NID2b, MSTM1a-MSTM1b (Jeffreys, Neumann et al. 2005). Cependant les distributions de CO laissent apparatre dans tous les cas des pics trs nettement spars les uns des autres. La rgion a1-yz1 chez le mas prsente en revanche une distribution de CO assez similaire celle du locus 130x (Yao et Schnable 2005)(figure 22). Il se pourrait donc que la rpartition relative des points chauds de recombinaison miotique obisse des contraintes spcifiques chez les plantes, diffrentes de celles des mammifres et des levures S. cerevisiae et S. pombe. Seule lidentification dun nombre plus important de points chauds de plantes permettra de rpondre cette question. Dans cette perspective, il est probable que le progrs des techniques de squenage massif permette de disposer prochainement de donnes trs nombreuses et prcises sur le dsquilibre de liaison lchelle du gnome entier chez A. thaliana. Cela pourrait aboutir lidentification de points chauds de recombinaison historiques analogues ceux dcrits chez lhomme. Ces points chauds historiques pourraient ensuite tre caractriss en dtails par lapproche de typage de pollen dcrite dans ce manuscrit. 108

3.5 Crossovers et non-crossovers au point chaud 130x dans le mutant Atmsh4


Dans le mutant Atmsh4, la frquence des CO observe au point chaud 130x est gale 5,2 % de celle du type sauvage. Ce rapport est lgrement infrieur ce qui a t calcul par la comparaison entre carte gntique et dcompte de foci MLH1 (7 %, voir section 3.3.), pour la moyenne du gnome, ce qui peut suggrer que la proportion de crossings-over de type II est variable lchelle du gnome, bien quil soit impossible de prciser dans quelle mesure cela se vrifierait une chelle sub-chromosomique. De faon assez surprenante, la distribution des points dchange dans Atmsh4 est notablement diffrente de celle du type sauvage (30 % des crossovers observs dans Atmsh4 sont situs en dehors de la distribution tablie dans le type sauvage). Etant donn que la distribution des CDB ne dpend a priori pas de MSH4, cela suggre une implication diffrente des deux voies de formation des CO au niveau de sites dinitiation de la recombinaison situs proximit les uns des autres. La raison dune telle diffrence est totalement inconnue. Contre toute attente, la frquence de conversion gnique est galement trs abaisse dans Atmsh4 (18 % de celle du type sauvage), mais dans une moindre mesure par rapport celle des CO. Cela semble impliquer que les intermdiaires prcoces de recombinaison (normalement destins former des CO de type I dpendants de MSH4) ne sont pas redirigs vers la voie de formation des NCO. En effet la frquence de conversion gnique naugmente pas en proportion de la diminution de frquence des CO. Il est possible que ces intermdiaires soient rpars en utilisant la chromatide sur. Cependant, puisque la frquence de conversion gnique diminue, il faut envisager que le gne AtMSH4 est galement ncessaire la formation dau moins une partie des NCO. MSH4 pourrait par exemple tre implique dans la stabilisation dun certain type dintermdiaires de recombinaison susceptible dengendrer des NCO, soit dans la voie de lHBDS (section 1.2.2.5.3.), soit dans la voie (hypothtique) de dissolution des dJH par un complexe RECQ4A-TOP3alpha-BLAP75 (section 1.2.2.5.1.). Ce rle nouveau est parfaitement compatible avec les observations cytologiques montrant que MSH4 est localise dans des foci trs nombreux ds le stade leptotne chez la souris (Kneitz, Cohen et al. 2000) et A. thaliana (Higgins, Armstrong et al. 2004). De plus, chez S. macrospora, Msh4 semble prsente tous les sites dinteraction entre homologues au dbut du zygotne, et savre ncessaire la progression normale de lappariement et du synapsis (Storlazzi, Gargano et al. 2010). Tous ces lments plaident en faveur dun rle de MSH4 beaucoup plus prcoce et gnral que les donnes accumules chez S. cerevisiae ne le laissaient supposer, la mutation de msh4 dans cette espce ne provoquant pas de 109

modification de la frquence des NCO. Cependant S. cerevisiae fournit quand mme un indice : dans un mutant msh4, la longueur des tracts de conversion associs aux CO est augmente, tandis que celle des tracts associs aux NCO est diminue (Mancera, Bourgon et al. 2008). Cette observation est compatible avec lhypothse selon laquelle il existerait non pas une mais deux voies de formation des NCO, dont lune seulement ferait intervenir MSH4. Il se pourrait alors que limportance de cette protine dans lune et lautre voie varie selon les espces, ce qui expliquerait les diffrences daltration de la frquence des NCO. Le rle dans la formation des NCO des autres gnes impliqus dans la diffrentiation des CO de type I chez A. thaliana (Mer3, Zyp1, Zip4, SHOC1, PTD) est actuellement inconnu. La fonction molculaire suppose de Mer3 (ADN hlicase) dsigne ce gne comme un des meilleurs candidats la caractrisation de NCO lchelle de points chauds en contexte mutant.

3.6 Bilan et perspectives


Trois conclusions majeures mergent de ce travail de thse : La taille des bivalents et linterfrence modulent conjointement le nombre et la distribution des crossovers le long du chromosome 4 dA. thaliana. Il existe des points chauds de recombinaison miotique, contenant des crossovers et des non crossovers, chez Arabidopsis thaliana. La distribution de ces points chauds est ellemme susceptible dobir des contraintes spcifiques aux plantes. MSH4 est impliqu dans la formation des non crossovers chez A. thaliana, et vraisemblablement chez tous les eucaryotes. Les programmes de recherche en cours dans notre quipe visent pour une part prciser et tendre la porte de ces rsultats : La distribution fine des CO lchelle du gnome entier dA. thaliana a t dtermine, par une approche similaire celle mise en uvre prcdemment pour le chromosome 4 (Drouaud, Mercier et al. 2007), et est actuellement analyse. De nouvelles rgions chaudes ont t identifies, au sein desquelles des points chauds pourront tre localiss. La caractrisation dvnements de NCO au point chaud 130x est poursuivie par Laurne Giraut. Loptimisation de la mthodologie permettra bientt de disposer dun trs grand nombre dvnements de conversion gnique non associs des CO, deux voire trois sites diffrents du point chaud.

110

Un second point chaud (14a) localis sur le bras court du chromosome 4 est actuellement tudi par H. Khademian dans le cadre de la prparation dune Thse de Doctorat. Des rsultats trs originaux, concernant notamment la variation de lactivit de recombinaison dans diffrents fonds gntiques hybrides (sauvages), sont attendus.

La distribution lchelle du gnome entier des vnement de CO et de NCO sera prochainement dtermine dans une srie de produits rciproques de miose (ttrades). Cela permettra (i) danalyser le dterminisme de la formation des CDBs initiatrices de la recombinaison et (ii) dclaircir le mcanisme de leur diffrentiation en CO et NCO, notamment en fonction de leur localisation chromosomique.

111

4 Matriels et mthodes
Les matriels et mthodes relatifs aux rsultats prsents dans cette thse sont inclus dans les articles, parus ou paratre, figurant dans la section 2. En outre, les protocoles de mise au point et de mise en uvre de la technique de typage de pollen font lobjet dun chapitre paratre dans un prochain volume de la srie Methods in Molecular Biology traitant de la recombinaison homologue, dit par H. Tsubouchi.

4.1 Characterization of meiotic crossovers in pollen from

Arabidopsis thaliana

112

Characterization of meiotic crossovers in pollen from Arabidopsis thaliana


Jan Drouaud* and Christine Mzard

Institut Jean-Pierre Bourgin UMR1318 INRA-AgroParisTech Institut National de la Recherche Agronomique, Centre de Versailles-Grignon Route de St-Cyr (RD10) 78026 Versailles Cedex France.
*

Corresponding author.

E-mail: jan.drouaud@versailles.inra.fr. Tel.: (33)130833851. Fax.: (33)130833319.

113

Abstract
Homologous recombination processes, which occur during the prophase of the first meiotic division, while generating new allelic combinations, are mechanistically important for the regular segregation of homologous chromosomes. They generate either crossovers, which are reciprocal exchanges between chromosome segments, or gene conversions. Both kinds of events occur in narrow regions (less than 10 kilobases) called hotspots, which are distributed along chromosomes. Classical genetic methods for CO characterization, which rely on the building of large populations and require appropriately located markers, are not well suited to the study of meiotic recombination hotspots. Here, we present a method based on allele-specific PCR amplification of single molecules from pollen genomic DNA. It allows detection, quantification and characterization of CO events arising at low frequencies in recombination hotspots.

Key words: Meiosis, crossover, pollen DNA, allele-specific PCR

114

1.

Introduction

During meiosis, the ploidy level is halved through a series of two cell divisions following a single round of DNA replication. The first division segregates homologous chromosomes, while the second division segregates sister chromatids, similar to mitosis. Hence, meiosis yields four cells each having half the number of chromosomes of the progenitor cell. The creation of physical connections between homologs is an absolute requirement for their proper segregation at the first division. In most species, this is achieved through the formation of crossovers (COs), which are reciprocal exchange of segments between homologous non-sister chromatids. COs result from the homologous repair of double-strand breaks (DSBs) formed early during the meiotic prophase. Other outcomes of that repair process are non crossovers (NCO, local non reciprocal exchange, sometimes called gene conversions) and sister chromatid exchanges (SCE) (1, 2). Besides their crucial mechanical role in the segregation of homologs, COs generate new allelic combinations. This provides both a substrate for natural selection, and the basis of genetic maps. It has been noticed from the very beginnings of genetics and cytogenetics that neither the rate of COs, nor their distribution along chromosomes, are uniform. These two features are controlled at several levels. First meiotic DSBs cluster in small regions (few kilobases) called hotspots, which are not homogenously distributed across chromosomes. Second the odds of a DSB to yield a CO or a NCO or a SCE, are likely intrinsically specific to each hotspot. Third the phenomenon of interference prevents COs from occurring close to each other (3). The distributions of COs at whole chromosome and genome levels have been extensively studied by analyzing the segregation of genetical markers in the offspring of hybrid parents. On the other hand cytological methods describe the distribution along chromosomes at the pachytene stage of meiotic prophase of structures, either electron-dense nodules or immunostained foci that correspond to COs (4). Recently, a completely new type of approach has arisen, which relies on the analysis of linkage desequilibrium (LD) among populations. This allows building haplotype maps of whole genomes, whose boundaries are thought to represent the preferential location of meiotic COs over evolutionary times. In human 50000 such historical hotspots have thus been described (5-8).

115

Such a global analysis of currently active meiotic hotspots is by no way feasible using the available tools of molecular biology. Nevertheless some have been extensively studied, either in bakers yeast or in mammals (mouse, human) (for review see (9, 10)). Among the tens of hotspots analyzed so far in higher eukaryotes, the reported frequency of COs does not exceed 1% of meioses. So it is clear that classical genetic analysis on siblings would need huge populations to get enough COs to characterize a hotspot accurately. On the other hand, cytological methods are not spatially accurate enough to describe hotspots. LD analysis methods cannot distinguish between actual and historical hotspots and they cannot detect evolutionary uprising hotspots, which have not yet significantly been involved in the reshuffling of haplotypes. Jeffreys and collaborators set up a technique called sperm typing that allows the recovery of CO molecules from sperm DNA (11). The procedure is based on a series of nested PCRs for isolating single recombinant molecules from pools of parental non-recombinant molecules. Several reviews have extensively described the use of this method in humans or mice(12, 13). Here, we will present an adaptation of this technique to the study of meiotic hotspots in DNA extracted from pollen grains, which are haploid structures producing sperm cells in higher plants. We will focus on the specificities of the design of allele-specific oligonucleotides that allow performing long PCR (up to 12 kb) on single molecules. We will also give protocols to perform such long PCR and to extract DNA from pollen grains. Using this pollen typing strategy, virtually any hotspot can be studied. Moreover, it can be adapted to the characterization of any kind of rare DNA variation in a population. The principle of this method is outlined in Figure 1. Briefly, single molecules, either parental or COs are detected in genomic DNA (gDNA) extracted from hybrid pollen grains, with two rounds of allele-specific PCR. This allows measuring the overall CO frequency in the studied region. Next, PCR amplified CO molecules are sequenced to map breakpoints, allowing the analysis of CO frequencies across the hotspot.

2.
2.1.
2.1.1.

Materials
DNA extraction
Extraction of DNA from pollen

1. 10% saccharose 2. Lysis buffer: 100 mM NaCl, 50 mM Tris-Hcl pH 8, 1 mM EDTA, 1% SDS. Add dithiothreitol (DTT) to 1 mM just prior to use. 3. Proteinase K 20 mg/ml. 116

4. Liquid phenol, saturated with 1M Tris-HCl pH 8. 5. Chloroform/isoamyl alcohol (IAA) (24/1 v/v). 6. 3 M sodium acetate pH 5.2. 7. Isopropanol. 8. 70 % ethanol. 9. RNaseA 10 mg/ml, DNase free. 10. TE buffer: 10 mM Tris-HCl pH 8, 1 mM EDTA.

2.1.2.

Extraction of DNA from leaves

11. Polyvinylpolypyrrolidone (PVPP). 12. Lysis buffer: 2% Cetyltrimethylammonium (CTAB), 100 mM Tris-HCl pH 8, 1.4 M NaCl, 20 mM EDTA. Add 2-mercaptoethanol to 10 mM just prior to use. 13. Chloroform/isoamyl alcohol (IAA) mix (24/1). 14. 3 M Sodium acetate pH 5.2. 15. Isopropanol. 16. 70 % ethanol. 17. TE buffer: 10 mM Tris-HCl pH 8, 1 mM EDTA.

2.1.3.

DNA purification

1. DNase free RNaseA 10 mg/ml. 2. Denaturation/binding buffer: 5M guanidine isothiocyanate, 50 mM Tris-HCl pH 8. 3. DNeasy Plant Mini Kit (Qiagen). 4. TE buffer: 10 mM Tris-HCl pH 8, 1 mM EDTA.

2.2.

Allele-Specific Long PCR


450 mM 110 mM 45 mM 67 mM 44 M 8 mM 8 mM 117

1. 10X PCR buffer, prepared as follows: Tris-HCl pH 8.8 (NH4)2SO4 MgCl2 2-mercaptoethanol EDTA dATP dCTP

2. Mix of Taq and Pfu DNA polymerases. See Note 2. 3. Desalted oligonucleotides. Stock solutions: 100 M in 5 mM Tris pH 8.8. 10X working solutions: 4 M in 5 mM Tris-HCl pH 8.8.

3.
3.1.
3.1.1.

Methods
Genomic DNA preparation
Extraction of genomic DNA from pollen

In the course of this procedure, pollen is first isolated from inflorescences. 1. Harvest Arabidopsis thaliana whole inflorescences in ice-cold 10 % saccharose. 2. Store at -20C or proceed directly to step 3. 3. Grind inflorescences in a minimal volume of 10% saccharose, using a Waring blendor. In most plant species, including A. thaliana, pollen wall is much more resistant to mechanical disruption than other tissues. This treatment bruises floral organs, so that intact pollen grains are released from anther locules. 4. Filter the homogenate through a 80 m mesh (nylon or steel). 5. Centrifuge the filtrate at 350 g for 10 min at 4C. 6. Discard supernatant. 7. Wash the pellet with ice-cold 10% saccharose. 8. Centrifuge the cell suspension at 100 g for 10 min at 4C. After this step, pollen grains are pelleted, while small cell and tissue fragments remain in the supernatant. 9. Discard the supernatant. 10. Repeat steps 7-9. 11. Store pollen at -20C or proceed directly to step 12. 118

12. Resuspend the cell pellet in 4 volumes of lysis buffer (see section 2.1.1). 13. Add proteinase K to 20 g/ml. 14. Incubate for 4 hours at 65C with occasional gentle homogenisation. 15. Add five to ten 2 mm diameter glass beads. 16. Vortex at full speed for 30 sec. 17. Add 1 l of the suspension into 10 l of water onto a microscope glass slide. Confirm the disruption of the cells with a microscope. See Note 3. 18. Proceed again to steps 16-17 until ~90% of pollen grains are disrupted. 19. In a chemical hood, add 1 volume of phenol saturated with 1M Tris-HCl pH 8. 20. Mix on a rocking wheel for 30 min. 21. Centrifuge at 15000 g for 10 min. 22. Transfer supernatant to a new tube, avoid pipetting any solid material. 23. In a chemical hood, add an equal volume of chloroform/IAA. Homogenate by gentle shaking. 24. Centrifuge at 15000 g for 10 min. 25. Transfer the supernatant to a new tube. 26. Add 0.7 volume of isopropanol. 27. Centrifuge at 15000 g for 10 min., at 4C. Discard the supernatant. 28. Wash the pellet with 1 ml of 70% ethanol. 29. Centrifuge at 15000 g for 2 min., at 4C. Discard the supernatant. Drain residual ethanol. 30. Let the pellet dry for 15 min at room temperature. 31. Dissolve the pellet in 100 l of TE buffer per gram of fresh material.

3.1.2.

Extraction of genomic DNA from leaves

Genomic DNA extracts from parents are used for testing the specificity of allele-specific oligonucleotides (ASOs) (sections 3.2.3.2. to 3.2.3.4.), while an extract from an F1 hybrid is used for testing their efficiency (section 3.2.3.5.) and performing control reactions (section 3.2.3.6.). 1. Pre-incubate 100 ml of lysis buffer at 65C. 2. Weigh an empty 50 ml Falcon type tube. Keep it on ice. 3. Harvest A. thaliana young rosettes leaves in the tube. 4. Weigh the tube again. Deduce the weight of fresh material. 119

5. Freeze leaves with liquid nitrogen in a mortar. 6. Add an equal amount of PVPP. 7. Grind with a pestle until getting a fine powder, taking care that the mortar keeps cold (add liquid nitrogen if necessary). 8. Grind again to get an even finer powder. 9. Transfer the powder to a 30 ml centrifuge tube and allow it to thaw to room temperature. 10. Add 5 ml of hot (65C) lysis buffer for each gram of fresh material and homogenate thoroughly. 11. Incubate for 30 min at 65C, with occasional gentle shaking. 12. Let the lysate cool to room temperature. 13. In a chemical hood, add an equal volume of chloroform/IAA. Homogenate by vigorous shaking. 14. Centrifuge at 15000 g for 10 min. 15. Transfer the supernatant to a new tube, avoid pipetting any solid material. 16. If needed centrifuge again and transfer the supernatant to a new tube. 17. Add 0.7 volume of isopropanol. 18. Centrifuge at 15000 g for 10 min, at 4C. Discard the supernatant. 19. Wash the pellet with 2 ml of 70% ethanol. 20. Centrifuge at 15000 g for 2 min, at 4C. Discard the supernatant. Drain residual ethanol. 21. Let the pellet dry for 15 min at room temperature. 22. Dissolve the pellet in 40 l of TE buffer per gram of fresh material. 23. Run 1 l of solution on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. Photograph the gel under UV, including a high exposure time in order to see faint bands (see Note 4). 24. Check DNA integrity. A noticeable smear is indicative of extensive DNA shearing. In such a case, DNA is expected to have a low amplifiability (see Section 3.1.4.) and should not be used for setting up pollen typing experiments, then proceed again to step 1.

3.1.3.

Purification of genomic DNA

1. Add 1/100 volume of 10 mg/ml RNase A to DNA samples to be purified. Incubate for 10 min at room temperature. 2. Add 4 volumes of binding buffer. 120

3. Pipet up to 650 l of mixture into the DNeasy mini spin column. 4. Proceed to step 14 of the Qiagen procedure (Mini protocol). 5. Adjust the volume of DNA solution to 40 l per gram of fresh material with TE buffer.

3.1.4.

Quantification of genomic DNA

During the processes of genomic DNA extraction and purification, a variable number of breaks intervene along chromosomes. Consequently, the proportion of molecules that can be actually PCR amplified (amplifiability) drops. Typically, it ranges from 20 to 50 % for amplicons longer than 8 kilobases. If high amounts of DNA are yielded, its mass concentration can be readily measured by spectrophotometry, or gel electrophoresis and ethidium bromide staining. In addition, the latter method allows monitoring the integrity of DNA, which is roughly indicative of its amplifiability. 1. Run 1 l of solution along with 100 ng, 200 ng, 400 ng and 800 ng of phage lambda DNA on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. 2. Photograph the gel under UV, including a high exposure time in order to see faint bands. Mass concentration is always an overestimate of the concentration of amplifiable molecules. Nevertheless, it can first be considered for carrying out the design procedure of oligonucleotides, as long as only one batch of gDNA is used. Ultimately, the concentration of amplifiable molecules will be determined by nested PCR amplification of single molecules, as described in section 3.3. Only this value should be considered for subsequent experiments. See Note 5.

3.2.

Designing and testing oligonucleotides

Two kinds of oligonucleotides are used for PCR experiments described below. Those which anneal to a site which is not specific to any haplotype (ie non polymorphic) will be referred to as Universal Oligonucleotides (UOs). On the other hand, some are intentionally positioned at polymorphic sites, so that they are intended to anneal specifically to DNA from one haplotype. The latter are coined Allele Specific Oligonucleotides (ASOs). See Figure 1A.

3.2.1.

Length of amplicons

The outcome of a PCR reaction performed on single molecules of template DNA depends primarily on the size of amplicons. We routinely amplify DNA fragments whose size reaches 10 kilobases (kb) from A. thaliana genomic DNA. Longer fragments can be obtained (up to 15 kb), but the consistency of results is expected to drop with amplicon size. 121

Since the distribution pattern of COs across the hotspot is generally unknown, the size of the region over which PCR reactions will be performed should not be purposely restricted. Ideally, PCR fragments should also encompass the hotspot flanking regions, which are devoid of COs. See Section 3.3.2.2.

3.2.2.

Designing the UOs

UOs are primarily intended to be used for testing ASOs. Two UOs are needed, which must be designed in the vicinity of the most outer ASOs (see Figure 1A). UO-ASO and ASO-ASO pairs will generate fragments of similar size, and thus are expected to perform similarly. This allows the indirect assessment of the quality of ASOs used for nested PCR amplification of long single molecule. UO_L is used for testing ASO_AR1, ASO_BR1, ASO_AR2 and ASO_BR2. UO_R is used for testing ASO_AL1, ASO_BL1, ASO_AL2 and ASO_BL2 (see Figure 1). UOs must anneal with gDNA at high temperatures (68C or above), so that ASOs with a lower Tm are the only limiting factor with respect to annealing with the DNA template. UOs must also be highly efficient, i.e. yield consistently high amounts of DNA. These two conditions should be assessed first using UO_L and UO_R together as described in section 3.2.3. Subsequently, UOs will be used in combination with candidate ASOs for determining their Topt, which is intended to be around 60C (see section 3.2.4.3), and for evaluating their efficiency (see section 3.2.4.4). The efficiency of the UOs can be assessed by performing a series of reactions with a decreasing amount of template gDNA. 1. Prepare a reaction pre-mix for 14 reactions as follows: 28 l of 4 M UO_L 28 l of 4 M UO_R 28 l of 10X PCR buffer 14 l of 0.5 U/l Taq:Pfu mix 196 l of H20 2. Combine 42 l of pre-mix and 2 l of 1.5 ng/l F1 leaf gDNA in PCR tube/well 1. 3. Add 12 l of H20 to the remaining pre-mix. Then aliquot 22 l into PCR tubes/wells 2 to 12. See Figure 2 for the rationale of serial dilution. 4. Transfer 22 l from tube/well 1 to tube/well 2. Mix. 5. Transfer 22 l from tube/well 2 to tube/well 3. Mix. So on until tube/well 12. gDNA is serially diluted at 1/2 from tube 1 to 12, starting from 1.5 ng.

122

10

6. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(68C;30 sec + 45 sec/kb)}x30(68C;90 sec/kb)(4C;). 7. Add 5 l of DNA loading dye, and run 10 l on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. Photograph the gel under UV, including a high exposure time in order to see faint bands. High efficiency UOs allow synthesizing an amount of DNA which can be visualized starting from as few as 6 pg (40 genomes, see Section 3.1.4.) of A. thaliana, which correspond to the 9th dilution. If UOs fail to produce this amount of DNA starting from 8 or 16 times more initial gDNA, a new combination must be tested until performing well.

3.2.3.

Designing the ASOs

The specificity of ASOs is the key to successful detection of CO molecules. Hence, while sometimes tedious, this step requires an extreme care.

3.2.3.1.

Choice of polymorphic sites

ASOs must be highly discriminative of parental polymorphic templates when annealing during PCR. It means that at a given temperature, a type A primer must anneal efficiently to type A genomic DNA, but not to type B. Therefore, genomic DNA regions where parental sequences are most dissimilar are favorite candidates for positioning ASOs. At candidate polymorphic sites, ASOs must be designed so that their 3 ends diverge as much as possible (see Figure 3A). Large insertions/deletions (INDELs) are the most favourable situation, provided they are not repeated. In the latter case, oligonucleotides encompassing the deletion should be absolutely avoided, because they always will anneal along their 3 end to DNA of the other type (see Figure 3B). Nonetheless, even if insertions are not repetitions, some caution must be taken when positioning ASOs across deletions, because looping of genomic DNA can cause non-specific annealing to occur. To avoid this, the 3end of the ASOs located beyond the site of insertion must be short, so that the hybridization of the 3 end of the ASO to genomic DNA of the other type will be destabilized by the adjacent loop (see Figure 3C). We currently limit the length of this 3 end to 6C equivalent (see Note 6). It happens most of time that only SNPs are available in the region of interest. Multiple SNPs close to each other are to be preferred over isolated ones. Only groups of SNPs less than 10 bases apart from each other should be considered as more interesting than isolated ones. 123

11

Nevertheless, isolated SNPs can sometimes prove to be sufficient for highly discriminative ASOs. Only preliminary set up experiments can provide such evidence.

3.2.3.2. Defining the optimal annealing temperature of ASOs


The length of the ASOs has to be chosen so that they anneal to their target site at a convenient temperature. Given that the annealing behaviour of an oligonucleotide depends heavily on PCR conditions, it cannot be predicted by dedicated softwares, which provide starting point temperatures only. Instead it should be empirically characterized by gradient PCR, considering the following rationale: Gradient PCR is informative about the less stable oligonucleotide used in the reaction: this is the ASO to be studied, while the other one is an UO purposely chosen to be very stable (annealing above 68C). Two informative temperatures can be determined from the amplification pattern along a gradient. Topt is the highest temperature for which the reaction yield reaches the maximum amount of product. Tmax is the highest temperature for which a product can be detected by gel electrophoresis. For every ASO, gradient PCR experiments must be performed in parallel with each type of parental gDNA, in order to define the range of temperature over which it anneals efficiently with its specific template only, i.e. between non specific Tmax and specific Topt (T, see Figure 4). Ideally, ASOs will anneal to its specific template only, even at the lower end of the gradient. Nevertheless, ASOs with T higher than 6C are also good candidates. Optimal results have been obtained in our laboratory for ASOs with specific Topt around 60C. 1. Prepare a reaction pre-mix for 28 reactions as follows: 56 l of 4 M UO 56 l of 4 M ASO 56 l of 10X PCR buffer 28 l of 0.5 U/l Taq:Pfu mix 378 l of H20 2. For each parent, combine 260 l of reaction pre-mix and 26 l of 1.5 ng/l leaf gDNA. 3. Aliquot 22 l of each mix into the wells of one plate row.

124

12

4. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(gradient from 56C to 68C;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 5. Add 5 l of DNA loading dye, and run 10 l on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. Photograph the gel under UV, including a high exposure time in order to see faint bands.

3.2.3.3. Maximizing the discrimination between polymorphic genomic targets


Whenever only moderately discriminative ASOs are found, their specificity can be improved by introducing additional mismatches close to their 3 end. This aims to decrease the stability of ASO/genomic DNA duplexes, but much more for the non-specific target than for the specific one. Such mismatches must be chosen carefully, neither too close to the 3 end, in order to keep annealing of the ASO to cognate template, nor too far, so as to decrease enough annealing of the ASO to non-cognate template. Fig 3 provides an example of successful designing such an extramismatch ASO, whereas the non-mismatch version was not discriminative enough. It must be noted that these mismatches decrease the annealing temperature of ASOs to specific gDNA sites. Consequently, nucleotides must be added at the 5 end of ASOs to compensate this lowering, and keep the melting temperature around the optimum (see Note 6).

3.2.3.4. Harmonizing the Topt of ASOs


The use of ASOs with different Topt in the same reaction should be avoided. More generally it seems desirable to harmonize Topt for all ASOs used in the analysis of one particular hotspot, as it does not preclude their use whatever their association in future experiments, either planned or not yet planned. This harmonization step involves a trial and error process: 1. Choose an ASO, taking the Tm predicted by your favourite primer design software as an estimation of its Topt. 2. Measure Topt by gradient PCR. 3. If necessary, add or remove nucleotides at the 5 end, in order to respectively increase or decrease the Tm. (see Note 6).

4. Proceed again to step 2.

125

13

3.2.3.5. Testing the efficiency of the ASOs


The efficiency requirement for ASOs is not as decisive as for UOs. Indeed, ASOs are used in nested PCR experiments. The yield of the second PCR is generally high enough for sensitive detection purposes, even if the efficiency of oligonucleotides is not tremendous. Moreover raising the number of PCR cycles can generally compensate for a moderate efficiency of ASOs. However, ASOs sometimes happen to perform very poorly so that their use should be avoided. Hence, the efficiency of ASOs should be evaluated, each in combination with a high efficiency UO (UOL with ASO_AR1, ASO_BR1, ASO_AR2 or ASO_BR2 - UOR with ASO_AL1, ASO_BL1, ASO_AL2 or ASO_BL2) using the procedure described in section 3.2.3. If necessary, the number of cycles can be adjusted in successive testing experiments. If an ASO fails to produce any detectable amount of DNA, starting from 50 pg (320 genomes, see Section 3.1.4.) of template or less, then its use for nested PCR amplification of single molecules is not recommended.

3.2.3.6. Testing the specificity of the ASOs: control reactions with somatic DNA
When amplifying single CO molecules by nested PCR, some undetectable non-specific products arise during the first reaction, as a consequence of misannealing of ASOs, most likely with DNA molecules of the other parental type at the homologous site, but also possibly at other genomic locations. They can nonetheless be abundant in terms of number of molecules. It may happen that these products are in turn amplified non-specifically during the second reaction, yielding this time a detectable product. In the course of the first PCR round, it may also happen that short fragments, either broken molecules initially present in the gDNA extract or truncated PCR products (resulting from the untimely termination of DNA synthesis, because of a break in template DNA for example), anneal with longer complementary fragments, next priming DNA synthesis. Whenever template and priming molecules are allelic, chimeric molecules will arise, which cannot be distinguished from true CO molecules. The occurrence of these both kinds of events must be evaluated by carrying out nested PCR control experiments using high amounts of leaf gDNA from F1 (AxB) hybrids, which is not intended to contain any CO molecule. All potential pairs of ASOs designed for the amplification of CO molecules should be tested: ASO_AL1/ASO_BR1 then ASO_AL2/ASO_BR2, and ASO_BL1/ASO_AR1 then ASO_BL2/ASO_AR2. 126

14

Given that artifactual products are likely to arise infrequently during the first PCR, hence stochastically, multiple (e.g. 6) identical and independent reactions are to be run in parallel. 1. Prepare a reaction pre-mix for 64 reactions as follows: 128 l of 4 M ASO_AL1 128 l of 4 M ASO_AR2 128 l of 10X PCR buffer 64 l of 0.5 U/l Taq:Pfu mix 755.2 l of H20 2. Combine 338.4 l of pre-mix and 57.6 l of 1.5 ng/l F1 leaf gDNA in tube 1. This is a mix for 18 reactions each containing 32000 genomes. 3. Add 147.2 l of H20 to the remaining pre-mix. Aliquot by 198 l in tubes 2 to 6. Each of these mixes contains 9 reactions without gDNA. 4. Transfer 198 l from tube 1 to tube 2. Mix. 5. So on until tube 6. gDNA is serially diluted at 1/2 from tube 1 to 6. 6. Aliquot 22 l of tube 6 into each well of column 6 of PCR plate 1. 7. Aliquot tube 5 by 22 l into column 5 of PCR plate 1. 8. So on until tube 1. 9. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 10. Dilute 1 l of PCR products in 50 l of (5 mM Tris-HCl pH 8, 0.01% Triton X-100). 11. Prepare a reaction pre-mix for 56 reactions as follows: 112 l of 4 M ASO_AL2 112 l of 4 M ASO_AR2 112 l of 10X PCR buffer 56 l of 0.5 U/l Taq:Pfu mix 784 l of H20 12. Aliquot 21 l of the pre-mix into the wells of PCR plate 2. 13. Transfer 1 l of diluted products from the first PCR into PCR plate 2. 14. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;).

127

15

15. Add 5 l of DNA loading dye, and run 10 l on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. Photograph the gel under UV, including a high exposure time, in order to see faint bands. The occurrence of faint bands for moderate to high amounts of DNA is not uncommon in the two first dilutions (16000 amplifiable genomes or more). This is 30 times the amount of pollen gDNA which is typically needed for the amplification of single CO molecules. Hence, it is not problematic as long as the actual CO rate is not exceedingly low. Of course, this is all the less worrying when the size of those faint bands is obviously different of that of specific products. Conversely, if products are detected for moderate to low amounts of DNA (i.e. at concentrations which are used for amplifying single CO molecules from F1 pollen gDNA), then one or several ASOs can be suspected to lack specificity, despite the outcome of gradient PCR analysis. In such cases, the exchange points between haplotypes are all expected to be located either before the second polymorphism, or after the penultimate one, the priming sites of left and right ASOs being defined respectively as the first and the last ones. This can be assessed readily by sequencing PCR products. If it turns out that ASOs are indeed not specific enough, they should not be used for pollen typing experiments.

3.3.

Amplification of single molecules.

The characterization and quantification of CO events in a gDNA extract relies on the specific amplification of single (or quasi single, see below) molecules, of either parental or recombinant (CO) type. In this way, two rounds of PCR, using nested sets of ASOs, are performed (see Figure 1A). Nested PCR serves two goals: A single round of PCR is not sufficient to get a detectable amount of product. Usually, two rounds yield a strong and consistent amplification. This allows detecting unambiguously all target molecules initially present in the reactions. The use of nested ASOs enhances the specificity of PCR amplification, because non-specific products from the first round of PCR, which are not expected to be abundant, are in turn unlikely to give rise to any detectable product during the second round of PCR. Single target molecules are isolated by dilution. When their average number per sample in a series of aliquots drops below 1, at least one reaction is expected to contain no template, and therefore to yield no PCR product. Actually, the proportion of these negative reactions can be estimated

128

16

using the Poisson law. Given m, the mean number of molecules per reaction, it provides the probability of a reaction to contain exactly k molecules: p(k) So: p(0) m0 e -m e -m 0! m k e -m k!

For example, if the mean number of molecules is 0.2, p(0) e -0.2 82% of reactions contain no molecule. Hence, the mean number of molecules per well, m, can be readily estimated from the proportion of negative wells P0 (which itself approximates p(0)): m ln(P0 ) In turn, m provides an estimate of the actual concentration of amplifiable molecules in the gDNA stock solution. See Figure 5 for an illustrated example. The variance of m can be calculated as described by (12).

3.3.1.

Quantification of gDNA by single-molecule PCR.

Given an uncharacterized gDNA extract, it is first necessary to perform nested PCR on a broad series of dilutions (e.g. 12), in order to get a rough, preliminary estimate of the concentration of amplifiable molecules. For each dilution a small number (e.g. 8) of aliquot reactions are carried out. For dilutions in which negative wells appear, Poisson formula allows calculating the concentration of molecules in the gDNA extract. 1. Prepare a reaction pre-mix for 115 reactions as follows: 230 l of 4 M ASO_AL1 230 l of 4 M ASO_AR1 230 l of 10X PCR buffer 115 l of 10 ng/l carrier DNA (see Note 7) 115 l of 0.5 U/l Taq:Pfu mix 1598.5 l of H20 2. Combine 262.8 l of pre-mix and 1.2 l of gDNA in tube 1. This is a mix for 12 reactions each containing 0.1 l of gDNA. 3. Add 10.3 l of H20 to the remaining pre-mix. Aliquot 198 l into tubes 2 to 12. 4. Transfer 66 l from tube 1 to tube 2. Mix. 129

17

5. Transfer 66 l from tube 2 to tube 3. Mix. 6. So on until tube 12. Amplifiable gDNA is serially diluted at 1/4 from tube 1 to 12. 7. Aliquot 22 l of tube 12 into the wells of column 12. 8. Aliquot 22 l of tube 11 into the wells of column 11. 9. So on until tube 1. 10. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 11. Dilute 1 l of PCR products in 50 l of (5 mM Tris-HCl pH 8, 0.01% Triton X-100). 12. Prepare a reaction pre-mix for 98 reactions as follows: 196 l of 4 M AL2 196 l of 4 M AR2 196 l of 10X PCR buffer 98 l of 0.5 U/l Taq:Pfu mix 1372 l of H20 13. Aliquot 21 l of the pre-mix into the wells of PCR plate 2. 14. Transfer 1 l of diluted products from the first PCR into PCR plate 2. 15. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 16. Add 5 l of DNA loading dye, and run 10 l on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. Photograph the gel under UV, including a high exposure time in order to see faint bands. 17. Assuming that the proportion of negative wells P0 in column i is the closest to 0.5, calculate the concentration of molecules in the gDNA extract as follows:

c 1 ln(P0 )

1 4 (i -1) 0.1

-ln(P0) is divided by the volume of DNA used in each reaction (0.1 l), then multiplied by a correction factor accounting for dilution in the ith column(4(i-1)). Next, carry out PCR upon a narrower range of 1/2 serial dilutions (e.g. 4), each with a higher number of aliquot reactions (e.g. 24), in order to get a more accurate estimation of c. 18. Prepare a reaction pre-mix for 128 reactions as follows: 256 l of 4 M ASO_AL1 256 l of 4 M ASO_AR1
130

18

19. Combine in tube 1: 1000 l of pre-mix and 100 l of gDNA diluted at 1/c1 in 5 ng/l carrier DNA. This is a mix for 50 reactions each expected to contain 2 amplifiable molecules. 20. Add 156 l of 5 ng/l carrier DNA to the remaining pre-mix. Aliquot 550 l in tubes 2 to 4. Each of these mixes contains 25 reactions without amplifiable gDNA. 21. Transfer 550 l from tube 1 to tube 2. Mix. So on until tube 4. 22. Amplifiable gDNA is serially diluted at 1/2 from tube 1 to 4. 23. Aliquot 22 l of tube 4 into the wells of rows G and H. So on until tube 1. 24. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 25. Dilute 1 l of PCR products in 50 l of (5 mM Tris-HCl pH 8, 0.01% Triton X-100). 26. Prepare a reaction pre-mix for 98 reactions as follows: 196 l of 4 M AL2 196 l of 4 M AR2 196 l of 10X PCR buffer 98 l of 0.5 U/l Taq:Pfu mix 1372 l of H20 27. Aliquot 21 l of the pre-mix into the wells of PCR plate 2. 28. Transfer 1 l of diluted products from the first PCR into PCR plate 2. 29. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 30. Add 5 l of DNA loading dye, and run 10 l on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. Photograph the gel under UV, including a high exposure time in order to see faint bands. 31. Assuming that the proportion of negative wells P0 in dilution j is the closest to 0.5, calculate the concentration of molecules in the gDNA extract as follows:

1 c 2 ln(P0 ) 2(j1) c 1 2 -ln(P0) is divided by the volume of DNA used in each reaction (2 l), then multiplied by a correction factor accounting for dilution in the jth column (2(j-1)xc1).

131

19

At last, 96 reactions are performed for one dilution only, providing a definitive estimate of target gDNA concentration in the stock solution. 32. Prepare a reaction mix for 98 reactions as follows: 196 l of 4 M ASO_AL1 196 l of 4 M ASO_AR1 196 l of 10X PCR buffer 98 l of 0.5 U/l Taq:Pfu mix 69.3 l of gDNA diluted at 1/c2 in 5 ng/l carrier DNA 126.7 l of 5 ng/l carrier DNA 1274 l of H20 33. Aliquot 22 l of the mix into the wells of PCR plate 1. 34. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 35. Dilute 1 l of PCR products in 50 l of (5 mM Tris-HCl pH8, 0.01% Triton X-100). 36. Prepare a reaction pre-mix for 98 reactions as follows: 196 l of 4 M AL2 196 l of 4 M AR2 196 l of 10X PCR buffer 98 l of 0.5 U/l Taq:Pfu mix 1372 l of H20 37. Aliquot 21 l of the pre-mix into the wells of PCR plate 2. 38. Transfer 1 l of diluted products from the first PCR into PCR plate 2. 39. Proceed to thermal cycling as follows: (92C;2 min){(92C;20 sec)(Topt;30 sec)(68C;45 sec/kb)}x30(68C;90 sec/kb)(4C;). 40. Add 5 l of DNA loading dye, and run 10 l on a 0.8 % agarose gel (containing 0.2 g/ml ethidium bromide) in 1X TBE. Photograph the gel under UV, including a high exposure time in order to see faint bands. 41. The proportion of negative wells is expected to be e-0.693=0.5. Given the real value P0, calculate the concentration of molecules in the gDNA extract as follows:
c3 ln(P0 ) c2 ln(0.5)

-ln(P0) is divided by the volume of DNA used in each reaction (-ln(0.5)=0.693 l), then multiplied by a correction factor accounting for dilution (c2).
132

20

Provided that very efficient ASOs have been used (see Section 3.2.5.1.), high PCR yields should be achieved. If it is not the case, the number of PCR cycles can be increased to 35, either for the first reaction, or for the second one, or for both. The last steps of this procedure (33-42) should next be carried out using sets of ASOs designed for detecting type B parental molecules: BL1/BR1 and BL2/BR2, for the first and second PCR respectively. Of course, results are expected to be the same than for type A parental molecules. The two values must next be summed, providing the total concentration of parental molecules in the hybrid.

3.3.2.

Amplification and characterization of single CO molecules.

3.3.2.1.

Quantification of CO molecules

Single CO molecules are intended to be amplified by two rounds of PCR, using a procedure similar to that of section 3.3.1. Assuming that control reactions performed with F1 somatic DNA (see Section 3.2.5.2.) allow a
priori ruling out the possibility that recombined molecules could be amplified non-specifically, i.e.

from parental type molecules, the outcome of single CO amplification experiments can be confidently envisioned. Nevertheless, during the first cycles of the first PCR round, parental molecules are present in very large excess over a single CO molecule to be amplified. Hence, they can possibly compete for annealing with ASOs. Consequently, the amplification efficiency of single CO molecules might be sub-optimal., yielding low amounts of DNA at the end of second PCR. When arising, this problem can possibly be solved by increasing the number of first PCR cycles. The method for quantifying CO molecules is identical to that for parental molecules, described in section 3.3.1., with the following modifications: ASO_AL1/ASO_BR1 next ASO_AL2/ASO_BR2 are used for detecting COs from A to B, and ASO_BL1/ASO_AR1 next ASO_BL2/ASO_AR2 are used for detecting COs from B to A (see Figure 1A). Carrier DNA is not required and can be replaced by H2O, since CO molecules are to be detected among a large excess of parental molecules. Whenever positive reactions can be easily distinguished from negative ones at some DNA dilution, they can be considered to arise most probably from CO molecules. Figure 6 displays a typical result of single CO molecules quantification.

133

21

Since the crossing over process always generates reciprocal exchanges, COs from A to B haplotype are expected to be as frequent as those from B to A. Then, once CO concentration has been determined for one orientation (e.g. A to B), only the last steps of the procedure (i.e. one dilution only, steps 33-42 in section 3.3.1.) are to be repeated for the other orientation (B to A). Once the concentration of CO molecules has been determined, it is divided by the concentration of parental molecules to get the CO frequency R. Note that, since a crossing over always generates two symmetrical molecules, CO rate is not the sum of A to B and B to A CO rates, but the average.

3.3.2.2. Characterization of CO molecules


Once CO rate has been measured, single CO events can be amplified in order to map CO breakpoints along the hotspot. This is most readily performed by sequencing PCR products. As discussed in section 3.3. positive wells among a series of aliquot reactions can result from the amplification of single or multiple molecules. The proportion of positive reactions issued from a single CO molecule can be expressed as a function of the proportion of negative wells: P0 ln(P0 ) P0 1 For example, if P0=80% of reactions are negative, ceeded from a single molecule. Hence, those PCR amplification products that are yielded by two CO molecules will generate two overlapping sequences over the region extending between the two CO breakpoints. As long as no INDEL type polymorphism is encountered, the mixed sequence will be readable, in particular at SNP positions where two overlapping peaks should be detected. In theory the two CO breakpoints can thus be mapped. In practice though, the analysis of mixed products often turns out unworkable, because of the occurrence of INDELs and/or the differential amplification of parental molecules. It should be noted that such mixed sequences cannot arise from the amplification of heteroduplex pollen DNA. Indeed, unlike spermatozoids in animals, pollen development includes mitotic divisions following meiosis. See Note 3. If ASOs are located far enough from the hotspot, no CO breakpoint should be detected in the most outer intervals of PCR products, that is between the first and second polymorphisms, or between the penultimate and last ones (the first and last polymorphisms are priming sites respectively for left and right inner ASOs). Whenever some events of this kind are observed, their status depends on the occurrence of COs in inner intervals:
134

0.8 ln(0.8) 89 % of positive reactions pro0.8 1

22

if the distribution is actually truncated, say on one side, then the origin of CO breakpoints in the corresponding outer interval cannot be ascertained. ASOs that are more external are required for amplifying the whole hotspot.

if most of outer COs breakpoints in inner intervals are located far away from the ASOs, then CO breakpoints in outer intervals should be suspected to arise from misannealing of ASOs, as discussed in section 3.2.5.2, and consequently discarded from the analysis.

3.3.2.3. Plotting CO rate distribution across the hotspot


Given: R N n1, n2, , nk l1, l2, , lk the CO rate over the whole hotspot the total number of mapped CO breakpoints the number of CO breakpoints mapped in intervals 1, 2, , k, respectively the size (in pb) of intervals 1, 2, , k, respectively

The CO rates (in cM/Mb) in intervals 1, 2, , k are respectively


n n n 1 R 10 2 2 R 10 2 k R 10 2 N N N , , , 10 -6 l 10 -6 l 10 -6 l 1 2 k

Results can be plotted as in Figure 1B.

4.

Notes
supplier, the yield of PCR reactions can vary over several orders of magnitude, especially for long amplicons. This is most likely due to a poisoning effect of Pfu by dUTP, which is generated at high temperature by dCTP deamination, but is also present as a trace contaminant in all commercial batches of dNTPs. BSA should be molecular biology grade (e.g. MPbiochemicals #BSAS2001). Depending on the supplier, slight variations of PCR yield can occur. A white precipitate sometimes appears in the 10X PCR buffer when preparing a new batch, or when thawing an aliquot. It must not be discarded. Instead, buffer should be carefully homogenized before aliquoting or using. Prior to any routine use, the performances of each new batch of buffer should be evaluated. For this purpose, control PCR assays are carried out using serial dilutions of a gDNA template, and their yield are compared to those obtained with the previous batch of buffer in the same conditions. See section 3.2.3. for setting up such an assay. If one gets only a moderate
135

1. Only ultra-pure dNTPs must be used (e.g. Roche 03622614001). Indeed, depending on the

23

2. Commercial blends of Taq and Pfu (or another Pyrococcus species) DNA polymerases are available that perform task very well. Alternatively, homemade mixes of enzymes can be used, but this requires setting up thorough procedures for quality control. Whatever the origin of polymerases, the amount to be used for getting a high yield of clean PCR product has to be determined for each batch. Carry out a series of PCR reactions with an increasing amount of polymerases mix, using universal oligonucleotides designed for setting up pollen typing experiments. Usually, a smear is observed for a given amount of enzyme and above. The optimum should be fixed at half of this quantity. 3. This harsh treatment is required for disrupting cells efficiently. In the same time, it breaks chromosomal DNA, thus lowering its amplifiability. It should then be used with much care. This method is highly efficient for breaking the wall of pollen grains, but is rather ineffective for immature gametophytes (microspores and young bicellular pollen). Hence gDNA is extracted mostly from tricellular pollen grains. 4. Since the extract contains mostly RNA, photometric quantification of DNA is impossible. In order to avoid trapping of ethidium bromide by RNA, add loading buffer supplemented with RNaseA (500 g/ml) to loaded samples. 5. 125 Mb is probably a gross underestimate of A. thaliana Col-0 genome size. Considering 147 Mb as a more plausible value, the weight of one haploid genome is 0.15 pg (14). 6. The contribution of individual nucleotides upon the Tm of an oligonucleotide cannot be accurately predicted, because it depends on many factors, including their sequence context, their position along the oligonucleotide. As a first approach, A or T nucleotides are considered to contribute 2C, and G or C 4C. These values are only indicative. In particular, nucleotides added at the 5 end of an oligonucleotide are expected to have a lesser effect upon its Tm. 7. At low DNA concentrations, a significant proportion of molecules are thought to adsorb onto plastic surfaces, where their availability as templates for polymerization becomes questionable. This concern is readily settled by adding carrier DNA (from calf thymus or salmon sperm) to the reaction.

Acknowledgements
We are grateful to Wayne Crismani, Mathilde Grelon, Anouchka Guyon, Arnaud Ronceret and Nathalie Vrielynck for critical reading of the manuscript and helpful comments. This work was supported by grants from INRA and ANR (COMEREC1 and COPATH)
136

24

References
1. 2. 3. 4. 5. Hunter, N. (2007) Meiotic recombination, in Molecular Genetics of Recombination (Aguilera, A. and Rothstein, R., eds.), Springer, Berlin, pp. 381-442. Whitby, M. C. (2005) Making crossovers during meiosis. Biochem Soc Trans. 33, 1451-5. Mezard, C., Vignard, J., Drouaud, J. and Mercier, R. (2007) The road to crossovers: plants have their say. Trends Genet. 23, 91-9. Anderson, L. K. and Stack, S. M. (2005) Recombination nodules in plants. Cytogenet Genome Res. 109, 198-204. Gabriel, S. B., Schaffner, S. F., Nguyen, H., Moore, J. M., Roy, J., Blumenstiel, B., Higgins, J., DeFelice, M., Lochner, A., Faggart, M., Liu-Cordero, S. N., Rotimi, C., Adeyemo, A., Cooper, R., Ward, R., Lander, E. S., Daly, M. J. and Altshuler, D. (2002) The structure of haplotype blocks in the human genome. Science. 296, 2225-9. McVean, G. A., Myers, S. R., Hunt, S., Deloukas, P., Bentley, D. R. and Donnelly, P. (2004) The fine-scale structure of recombination rate variation in the human genome. Science. 304, 581-4. HapMap (2005) A haplotype map of the human genome. Nature. 437, 1299-320. Myers, S., Bottolo, L., Freeman, C., McVean, G. and Donnelly, P. (2005) A fine-scale map of recombination rates and hotspots across the human genome. Science. 310, 321-4. Petes, T. D. (2001) Meiotic recombination hot spots and cold spots. Nat Rev Genet. 2, 360-9. Kauppi, L., Jeffreys, A. J. and Keeney, S. (2004) Where the crossovers are: recombination distributions in mammals. Nat Rev Genet. 5, 413-24. Jeffreys, A. J., Murray, J. and Neumann, R. (1998) High-resolution mapping of crossovers in human sperm defines a minisatellite-associated recombination hotspot. Mol Cell. 2, 26773. Baudat, F. and de Massy, B. (2009) Parallel detection of crossovers and noncrossovers in mouse germ cells, in Methods Mol Biol eds.), pp.305-22 Kauppi, L., May, C. A. and Jeffreys, A. J. (2009) Analysis of meiotic recombination products from human sperm, in Methods Mol Biol eds.), pp.323-55 Bennett, M. D., Leitch, I. J., Price, H. J. and Johnston, J. S. (2003) Comparisons with Caenorhabditis (approximately 100 Mb) and Drosophila (approximately 175 Mb) using flow cytometry show genome size in Arabidopsis to be approximately 157 Mb and thus approximately 25% larger than the Arabidopsis genome initiative estimate of approximately 125 Mb. Ann Bot. 91, 547-57.

6.

7. 8. 9. 10. 11.

12. 13. 14.

137

25

Figure legends
Figure 1. Outline of the pollen typing method.

(A) Polymorphisms in parental A (white) and B (black) sequences are represented by circles. A and B-type Allele-specific oligonucleotides (ASOs) are depicted respectively as white and black triangles. Universal oligonucleotides (UOs) are displayed as gray triangles. (B) Procedure for PCR amplification and mapping of CO events.

Figure 2. Making serial dilutions.

1 2 3 4 5

Prepare a mix for (8+1)x(12+1)+1=118 reactions without template DNA. Add 2x(8+1)=18 reactions into tube 1. Add DNA. Add water to final volume. Add water to final volume into the remaining mix, which contains 118-18=100 reactions. Aliquot 8+1=9 reactions into tubes 2 to 12. Transfer 8+1=9 reactions from tube 1 into tube 2. Mix. Transfer 8+1=9 reactions from tube 2 into tube 3. Mix. So on until tube 12.

Aliquot tube 1 into column 1. Aliquot tube 2 into column 2. So on until tube 12.

Figure 3. Sample cases for designing ASOs.

Aligned A and B parental sequences are represented by gray and black horizontal lines, respectively. Identities are represented by vertical plain lines and SNPs by vertical dotted lines. Aspecific candidate ASOs are represented by arrowed lines. Insertions in B with respect to A are represented as a loop. Direct repeats in B sequence are indicated as thin arrowed lines.

Figure 4. Sample cases of ASO testing by gradient PCR.

Left part: ASOs #1 and #2 sequences aligned to parental sequences. Mismatched nucleotide in ASO #2 is highlighted. Right part: Panels 1 and 3: PCR is performed on gDNA from parent A. Panels 2 and 4: PCR is performed on gDNA from parent B.
138

26

Panels 1 and 2: ASO #1. The difference between specific Topt and non-specific Tmax (T) is 4.6C. The ASO is poorly specific. Panels 3 and 4: ASO #2. Extra mismatch in ASO sequence increasesT to more than 10C. This version of the ASO is highly specific.

Figure 5. Quantification of Poisson distributed molecules.

A genomic DNA extract is diluted 1/10000 (A) or 1/20000 (B). Next 96 reactions are performed, each with 1 l of diluted DNA. The numbers of wells containing either 0, 1, 2, 3 or 4 molecules follow a Poisson distribution. They are indicated in the first line of each panel and displayed above as grey bars. The corresponding numbers of molecules are shown in the last line. The real value of average molecules number per well m is very close to its theoretical value which is calculated as -ln(P0). In that case, the estimated concentration of amplifiable DNA molecules is 0.453x20000~0.901x10000~9039 molecules/l, which corresponds to 1.36 ng/l.

Figure 6. Sample PCR amplification of single CO molecules.

Each reaction has been performed using 1 l of pollen gDNA, diluted at 1/64. Stock solution contains 32550 parental molecules per l. Among these 46 reactions, 14 are positive and 32 are negative. The estimated mean number of CO molecules per reaction is -ln(32/46)=0.363. The estimated total number of CO molecules is 0.363x46=16.7. The concentration of CO molecules in the gDNA extract is 0.363x64=23.2 CO/l. CO rate is 23.2/32550=1/350

139

27

Figures
Figure 1

140

28

Figure 2

141

29

Figure 3

142

30

Figure 4

143

31

Figure 5

144

32

Figure 6

145

33

5 Rfrences bibliographiques
1 Abirached-Darmency, M., D. Zickler, et al. (1983). "Synaptonemal complex et recombination nodules in rye (Secale cereale)." Chromosoma 88(4): 299306. Agarwal, S. et G. S. Roeder (2000). "Zip3 provides a link between recombination enzymes et synaptonemal complex proteins." Cell 102(2): 245-55. Albini, S. M. et G. H. Jones (1987). "Synaptonemal complex spreading in Allium cepa et A. fistulosum. I. The initiation et sequence of pairing." Chromosoma 95(5): 324-338. Allers, T. et M. Lichten (2001). "Differential timing et control of noncrossover et crossover recombination during meiosis." Cell 106(1): 4757. Allers, T. et M. Lichten (2001). "Intermediates of yeast meiotic recombination contain heteroduplex DNA." Mol Cell 8(1): 225-31. Alpi, A., P. Pasierbek, et al. (2003). "Genetic et cytological characterization of the recombination protein RAD-51 in Caenorhabditis elegans." Chromosoma 112(1): 6-16. Andersen, S. L., D. T. Bergstralh, et al. (2009). "Drosophila MUS312 et the vertebrate ortholog BTBD12 interact with DNA structure-specific endonucleases in DNA repair et recombination." Mol Cell 35(1): 128-35. Anderson, L. K., G. G. Doyle, et al. (2003). "High-resolution crossover maps for each bivalent of Zea mays using recombination nodules." Genetics 165(2): 849-65. Anderson, L. K., K. D. Hooker, et al. (2001). "The distribution of early recombination nodules on zygotene bivalents from plants." Genetics 159(3): 1259-69. Anderson, L. K., H. H. Offenberg, et al. (1997). "RecA-like proteins are components of early meiotic nodules in lily." Proc Natl Acad Sci U S A 94(13): 6868-73. Anderson, L. K., A. Reeves, et al. (1999). "Distribution of crossing over on mouse synaptonemal complexes using immunofluorescent localization of MLH1 protein." Genetics 151(4): 1569-79. Anderson, L. K. et S. M. Stack (2005). "Recombination nodules in plants." Cytogenet Genome Res 109(1-3): 198-204. Anderson, L. K., S. M. Stack, et al. (1988). "Spreading synaptonemal complexes from Zea mays. I. No synaptic adjustment of inversion loops during pachytene." Chromosoma 96(4): 295-305.
146

5 6

10

11

12 13

14

Argueso, J. L., A. W. Kijas, et al. (2003). "Systematic mutagenesis of the Saccharomyces cerevisiae MLH1 gene reveals distinct roles for Mlh1p in meiotic crossing over et in vegetative et meiotic mismatch repair." Mol Cell Biol 23(3): 873-86. Argueso, J. L., J. Wanat, et al. (2004). "Competing crossover pathways act during meiosis in Saccharomyces cerevisiae." Genetics 168(4): 1805-16. Armstrong, S. J., A. P. Caryl, et al. (2002). "Asy1, a protein required for meiotic chromosome synapsis, localizes to axis-associated chromatin in Arabidopsis et Brassica." J Cell Sci 115(Pt 18): 3645-55. Armstrong, S. J., F. C. Franklin, et al. (2001). "Nucleolus-associated telomere clustering et pairing precede meiotic chromosome synapsis in Arabidopsis thaliana." J Cell Sci 114(Pt 23): 4207-17. Armstrong, S. J., F. C. H. Franklin, et al. (2003). "A meiotic time-course for Arabidopsis thaliana." Sexual Plant Reproduction 16(3): 141-149. Ashley, T. (2008). Synaptic et Recombination Nodules in Mammals: Structural Continuity with Shifting Protein Composition. Recombination et Meiosis - Crossing-Over et Disjunction: 153-194. Bachrati, C. Z., R. H. Borts, et al. (2006). "Mobile D-loops are a preferred substrate for the Bloom's syndrome helicase." Nucleic Acids Res 34(8): 2269-79. Banerjee, R. et G. H. Jones (1999). "Initiation et progression of homologous chromosome synapsis in Crepis capillaris: Variations on a theme." Genome 42(5): 867-73. Bannister, L. A. et J. C. Schimenti (2004). "Homologous recombinational repair proteins in mouse meiosis." Cytogenet Genome Res 107(3-4): 191200. Barber, L. J., J. L. Youds, et al. (2008). "RTEL1 maintains genomic stability by suppressing homologous recombination." Cell 135(2): 261-71. Barlow, A. L., F. E. Benson, et al. (1997). "Distribution of the Rad51 recombinase in human et mouse spermatocytes." Embo J 16(17): 5207-15. Barlow, A. L. et M. A. Hulten (1998). "Crossing over analysis at pachytene in man." Eur J Hum Genet 6(4): 350-8. Basheva, E. A., C. J. Bidau, et al. (2008). "General pattern of meiotic recombination in male dogs estimated by MLH1 et RAD51 immunolocalization." Chromosome Res 16(5): 709-19. Baudat, F., J. Buard, et al. (2010). "PRDM9 is a major determinant of meiotic recombination hotspots in humans et mice." Science 327(5967): 836-40. Baudat, F. et B. de Massy (2007). "Cis- et Trans-Acting Elements Regulate the Mouse Psmb9 Meiotic Recombination Hotspot." PLoS Genet 3(6): e100.
147

15 16

17

18 19

20

21

22

23 24 25 26

27 28

29

Baudat, F., K. Manova, et al. (2000). "Chromosome synapsis defects et sexually dimorphic meiotic progression in mice lacking Spo11." Mol Cell 6(5): 989-98. Baudat, F. et A. Nicolas (1997). "Clustering of meiotic double-strand breaks on yeast chromosome III." Proc Natl Acad Sci U S A 94(10): 5213-8. Bergerat, A., B. de Massy, et al. (1997). "An atypical topoisomerase II from Archaea with implications for meiotic recombination." Nature 386(6623): 414-7. Beye, M., I. Gattermeier, et al. (2006). "Exceptionally high levels of recombination across the honey bee genome." Genome Res 16(11): 1339-44. Birdsell, J. A. (2002). "Integrating genomics, bioinformatics, et classical genetics to study the effects of recombination on genome evolution." Mol Biol Evol 19(7): 1181-97. Bleuyard, J. Y., M. E. Gallego, et al. (2005). "Differing requirements for the Arabidopsis Rad51 paralogs in meiosis et DNA repair." Plant J 41(4): 533-45. Bleuyard, J. Y. et C. I. White (2004). "The Arabidopsis homologue of Xrcc3 plays an essential role in meiosis." Embo J 23(2): 439-49. Blitzblau, H. G., G. W. Bell, et al. (2007). "Mapping of meiotic singlestranded DNA reveals double-stranded-break hotspots near centromeres et telomeres." Curr Biol 17(23): 2003-12. Boehnke, M., N. Arnheim, et al. (1989). "Fine-structure genetic mapping of human chromosomes using the polymerase chain reaction on single sperm: experimental design considerations." Am J Hum Genet 45(1): 21-32. Bois, P. R. (2007). "A highly polymorphic meiotic recombination mouse hot spot exhibits incomplete repair." Mol Cell Biol 27(20): 7053-62. Bolcun-Filas, E., Y. Costa, et al. (2007). "SYCE2 is required for synaptonemal complex assembly, double strand break repair, et homologous recombination." J Cell Biol 176(6): 741-7. Bolcun-Filas, E., R. Speed, et al. (2009). "Mutation of the Mouse Syce1 Gene Disrupts Synapsis et Suggests a Link between Synaptonemal Complex Structural Components et DNA Repair." PLoS Genet 5(2): e1000393. Borde, V., N. Robine, et al. (2009). "Histone H3 lysine 4 trimethylation marks meiotic recombination initiation sites." Embo J 28(2): 99-111. Borner, G. V., N. Kleckner, et al. (2004). "Crossover/noncrossover differentiation, synaptonemal complex formation, et regulatory surveillance at the leptotene/zygotene transition of meiosis." Cell 117(1): 29-45. Borodin, P. M., E. A. Basheva, et al. (2009). "Immunocytological analysis of meiotic recombination in the American mink (Mustela vison)." Anim Genet 40(2): 235-8.
148

30 31

32 33

34 35 36

37

38 39

40

41 42

43

44

Borodin, P. M., T. V. Karamysheva, et al. (2008). "Recombination map of the common shrew, Sorex araneus (Eulipotyphla, Mammalia)." Genetics 178(2): 621-32. Borts, R. H. et J. E. Haber (1989). "Length et distribution of meiotic gene conversion tracts et crossovers in Saccharomyces cerevisiae." Genetics 123(1): 69-80. Boulton, A., R. S. Myers, et al. (1997). "The hotspot conversion paradox et the evolution of meiotic recombination." Proc Natl Acad Sci U S A 94(15): 8058-63. Brar, G. A., A. Hochwagen, et al. (2009). "The multiple roles of cohesin in meiotic chromosome morphogenesis et pairing." Mol Biol Cell 20(3): 103047. Bridges, C. B. (1935). "Salivary chromosome maps with a key to the banding of the chromosomes of Drosophila melanogaster." J Hered 26(2): 60-64. Broman, K. W., L. B. Rowe, et al. (2002). "Crossover interference in the mouse." Genetics 160(3): 1123-31. Broman, K. W. et J. L. Weber (2000). "Characterization of human crossover interference." Am J Hum Genet 66(6): 1911-26. Brown, P. W., L. Judis, et al. (2005). "Meiotic synapsis proceeds from a limited number of subtelomeric sites in the human male." Am J Hum Genet 77(4): 556-66. Buard, J., P. Barthes, et al. (2009). "Distinct histone modifications define initiation et repair of meiotic recombination in the mouse." Embo J. Buhler, C., V. Borde, et al. (2007). "Mapping Meiotic Single-Strand DNA Reveals a New Landscape of DNA Double-Strand Breaks in Saccharomyces cerevisiae." PLoS Biol 5(12): e324. Burnham, C. R., J. T. Stout, et al. (1972). "Chromosome Pairing in Maize." Genetics 71(1): 111-126. Burt, A., G. Bell, et al. (1991). "Sex differences in recombination." J Evol Biol 4: 259-77. Busygina, V., M. G. Sehorn, et al. (2008). "Hed1 regulates Rad51-mediated recombination via a novel mechanism." Genes Dev 22(6): 786-95. Cai, X., F. Dong, et al. (2003). "The Arabidopsis SYN1 cohesin protein is required for sister chromatid arm cohesion et homologous chromosome pairing." J Cell Sci 116(Pt 14): 2999-3007. Cao, L., E. Alani, et al. (1990). "A pathway for generation et processing of double-strand breaks during meiotic recombination in S. cerevisiae." Cell 61(6): 1089-101.

45

46

47

48 49 50 51

52 53

54 55 56 57

58

149

59

Carmi, P., P. Holm, et al. (1978). "The pachytene karyotype of Schizophyllum commune analyzed by three dimensional reconstruction of synaptonemal complexes." Carlsberg Research Communications 43(2): 117132. Caryl, A. P., S. J. Armstrong, et al. (2000). "A homologue of the yeast HOP1 gene is inactivated in the Arabidopsis meiotic mutant asy1." Chromosoma 109(1-2): 62-71. Celerin, M., S. T. Merino, et al. (2000). "Multiple roles of Spo11 in meiotic chromosome behavior." Embo J 19(11): 2739-50. Cha, R. S., B. M. Weiner, et al. (2000). "Progression of meiotic DNA replication is modulated by interchromosomal interaction proteins, negatively by Spo11p et positively by Rec8p." Genes Dev 14(4): 493-503. Chakravarti, A., K. H. Buetow, et al. (1984). "Nonuniform recombination within the human beta-globin gene cluster." Am J Hum Genet 36(6): 123958. Chakravarti, A., S. C. Elbein, et al. (1986). "Evidence for increased recombination near the human insulin gene: implication for disease association studies." Proc Natl Acad Sci U S A 83(4): 1045-9. Chelysheva, L., S. Diallo, et al. (2005). "AtREC8 et AtSCC3 are essential to the monopolar orientation of the kinetochores during meiosis." J Cell Sci 118(Pt 20): 4621-32. Chelysheva, L., G. Gendrot, et al. (2007). "Zip4/Spo22 is required for class I CO formation but not for synapsis completion in Arabidopsis thaliana." PLoS Genet 3(5): e83. Chelysheva, L., D. Vezon, et al. (2008). "The Arabidopsis BLAP75/Rmi1 homologue plays crucial roles in meiotic double-strand break repair." PLoS Genet 4(12): e1000309. Chen, S. Y., T. Tsubouchi, et al. (2008). "Global analysis of the meiotic crossover landscape." Dev Cell 15(3): 401-15. Chen, Y. K., C. H. Leng, et al. (2004). "Heterodimeric complexes of Hop2 et Mnd1 function with Dmc1 to promote meiotic homolog juxtaposition et strand assimilation." Proc Natl Acad Sci U S A 101(29): 10572-7. Chi, P., J. San Filippo, et al. (2007). "Bipartite stimulatory action of the Hop2-Mnd1 complex on the Rad51 recombinase." Genes Dev 21(14): 174757. Cobbs, G. (1978). "Renewal process approach to the theory of genetic linkage: case of no chromatid interference." Genetics 89(3): 563-581.

60

61 62

63

64

65

66

67

68 69

70

71

150

72

Codina-Pascual, M., M. Campillo, et al. (2006). "Crossover frequency et synaptonemal complex length: their variability et effects on human male meiosis." Mol Hum Reprod 12(2): 123-33. Colaiacovo, M. P., A. J. MacQueen, et al. (2003). "Synaptonemal complex assembly in C. elegans is dispensable for loading strand-exchange proteins but critical for proper completion of recombination." Dev Cell 5(3): 463-74. Coop, G. et S. R. Myers (2007). "Live Hot, Die Young: Transmission Distortion in Recombination Hotspots." PLoS Genet 3(3): e35. Coop, G., X. Wen, et al. (2008). "High-resolution mapping of crossovers reveals extensive variation in fine-scale recombination patterns among humans." Science 319(5868): 1395-8. Corredor, E., A. J. Lukaszewski, et al. (2007). "Terminal regions of wheat chromosomes select their pairing partners in meiosis." Genetics 177(2): 699706. Couteau, F., F. Belzile, et al. (1999). "Random chromosome segregation without meiotic arrest in both male et female meiocytes of a dmc1 mutant of Arabidopsis." Plant Cell 11(9): 1623-34. Couteau, F., W. Goodyer, et al. (2004). "Finding et keeping your partner during meiosis." Cell Cycle 3(8): 1014-6. Couteau, F. et M. Zetka (2005). "HTP-1 coordinates synaptonemal complex assembly with homolog alignment during meiosis in C. elegans." Genes Dev 19(22): 2744-56. Cromie, G. A., R. W. Hyppa, et al. (2007). "A discrete class of intergenic DNA dictates meiotic DNA break hotspots in fission yeast." PLoS Genet 3(8): e141. Cromie, G. A., R. W. Hyppa, et al. (2006). "Single Holliday junctions are intermediates of meiotic recombination." Cell 127(6): 1167-78. Cullen, M., S. P. Perfetto, et al. (2002). "High-resolution patterns of meiotic recombination across the human major histocompatibility complex." Am J Hum Genet 71(4): 759-76. Darlington, C. D. (1931). "Meiosis." Biological Reviews 6(3): 221-264. Darlington, C. D. (1935). "The time, place et action of crossing-over." Journal of Genetics 31(2): 185-212. Dawe, R. K., J. W. Sedat, et al. (1994). "Meiotic chromosome pairing in maize is associated with a novel chromatin organization." Cell 76(5): 901-12. de Boer, E., A. J. Dietrich, et al. (2007). "Meiotic interference among MLH1 foci requires neither an intact axial element structure nor full synapsis." J Cell Sci 120(Pt 5): 731-6.

73

74 75

76

77

78 79

80

81 82

83 84 85 86

151

87 88

de Boer, E., P. Stam, et al. (2006). "Two levels of interference in mouse meiotic recombination." Proc Natl Acad Sci U S A 103(25): 9607-12. Dernburg, A. F., K. McDonald, et al. (1998). "Meiotic recombination in C. elegans initiates by a conserved mechanism et is dispensable for homologous chromosome synapsis." Cell 94(3): 387-98. Detloff, P., M. A. White, et al. (1992). "Analysis of a gene conversion gradient at the HIS4 locus in Saccharomyces cerevisiae." Genetics 132(1): 113-23. Dherin, C., E. Gueneau, et al. (2009). "Characterization of a highly conserved binding site of Mlh1 required for exonuclease I-dependent mismatch repair." Mol Cell Biol 29(3): 907-18. Dooner, H. K. (1986). "Genetic Fine Structure of the BRONZE Locus in Maize." Genetics 113(4): 1021-1036. Dooner, H. K. (1998). "Reply: On the Possible Occurrence of Conversion Polarity at the bronze Locus." Plant Cell 10(5): 646-648. Dooner, H. K. (2002). "Extensive interallelic polymorphisms drive meiotic recombination into a crossover pathway." Plant Cell 14(5): 1173-83. Dooner, H. K. et I. M. Martinez-Ferez (1997). "Recombination occurs uniformly within the bronze gene, a meiotic recombination hotspot in the maize genome." Plant Cell 9(9): 1633-1646. Dray, E., N. Siaud, et al. (2006). "Interaction between Arabidopsis Brca2 et its partners Rad51, Dmc1, et Dss1." Plant Physiol 140(3): 1059-69. Dresser, M. E. et C. N. Giroux (1988). "Meiotic chromosome behavior in spread preparations of yeast." J Cell Biol 106(3): 567-73. Dreszer, T. R., G. D. Wall, et al. (2007). "Biased clustered substitutions in the human genome: the footprints of male-driven biased gene conversion." Genome Res 17(10): 1420-30. Drouaud, J., C. Camilleri, et al. (2006). "Variation in crossing-over rates across chromosome 4 of Arabidopsis thaliana reveals the presence of meiotic recombination "hot spots"." Genome Res 16(1): 106-14. Drouaud, J., R. Mercier, et al. (2007). "Sex-Specific Crossover Distributions et Variations in Interference Level along Arabidopsis thaliana Chromosome 4." PLoS Genet 3(6): e106. Dupaigne, P., C. Le Breton, et al. (2008). "The Srs2 helicase activity is stimulated by Rad51 filaments on dsDNA: implications for crossover incidence during mitotic recombination." Mol Cell 29(2): 243-54. Duret, L. et P. F. Arndt (2008). "The impact of recombination on nucleotide substitutions in the human genome." PLoS Genet 4(5): e1000071.

89

90

91 92 93 94

95 96 97

98

99

100

101

152

102 103 104

Duret, L. et N. Galtier (2009). "Biased Gene Conversion et the Evolution of Mammalian Genomic Landscapes." Annu Rev Genomics Hum Genet. Egel, R. (1978). "Synaptonemal complex et crossing-over: structural support or interference?" Heredity 41(2): 233-37. Ehmsen, K. et W.-D. Heyer (2008). Biochemistry of Meiotic Recombination: Formation, Processing, et Resolution of Recombination Intermediates. Recombination et Meiosis - Models, Means, et Evolution: 91-164. Falque, M., L. K. Anderson, et al. (2009). "Two Types of Meiotic Crossovers Coexist in Maize." Plant Cell. Falque, M., R. Mercier, et al. (2007). "Patterns of recombination et MLH1 foci density along mouse chromosomes: modeling effects of interference et obligate chiasma." Genetics 176(3): 1453-67. Fan, Q. Q., F. Xu, et al. (1997). "Competition between adjacent meiotic recombination hotspots in the yeast Saccharomyces cerevisiae." Genetics 145(3): 661-70. Fearnhead, P., R. M. Harding, et al. (2004). "Application of coalescent methods to reveal fine-scale rate variation et recombination hotspots." Genetics 167(4): 2067-81. Ferguson, K. A., S. Leung, et al. (2009). "Distribution of MLH1 foci et interfocal distances in spermatocytes of infertile men." Hum Reprod. Foss, E., R. Lande, et al. (1993). "Chiasma interference as a function of genetic distance." Genetics 133(3): 681-91. Foss, E. J. et F. W. Stahl (1995). "A test of a counting model for chiasma interference." Genetics 139(3): 1201-9. Fox, D. P. (1973). "The control of chiasma distribution in the locust, Schistocerca gregaria (Forskl)." Chromosoma 43(3): 289-328. Franklin, F. C., J. D. Higgins, et al. (2006). "Control of meiotic recombination in Arabidopsis: role of the MutL et MutS homologues." Biochem Soc Trans 34(Pt 4): 542-4. Froenicke, L., L. K. Anderson, et al. (2002). "Male mouse recombination maps for each autosome identified by chromosome painting." Am J Hum Genet 71(6): 1353-68. Fukushima, K., Y. Tanaka, et al. (2000). "Dmc1 of Schizosaccharomyces pombe plays a role in meiotic recombination." Nucleic Acids Res 28(14): 2709-16. Fung, J. C., B. Rockmill, et al. (2004). "Imposition of crossover interference through the nonrandom distribution of synapsis initiation complexes." Cell 116(6): 795-802.

105 106

107

108

109 110 111 112 113

114

115

116

153

117

Gabriel-Robez, O., C. Ratomponirina, et al. (1986). "Synapsis et synaptic adjustment in an infertile human male heterozygous for a pericentric inversion in chromosome 1." Hum Genet 72(2): 148-52. Game, J. C., K. C. Sitney, et al. (1989). "Use of a ring chromosome et pulsedfield gels to study interhomolog recombination, double-strand DNA breaks et sister-chromatid exchange in yeast." Genetics 123(4): 695-713. Gaskell, L. J., F. Osman, et al. (2007). "Mus81 cleavage of Holliday junctions: a failsafe for processing meiotic recombination intermediates?" Embo J 26(7): 1891-901. Gaut, B. S., S. I. Wright, et al. (2007). "Recombination: an underappreciated factor in the evolution of plant genomes." Nat Rev Genet 8(1): 77-84. Gerton, J. L., J. DeRisi, et al. (2000). "Global mapping of meiotic recombination hotspots et coldspots in the yeast Saccharomyces cerevisiae." Proc Natl Acad Sci U S A 97(21): 11383-90. Gillies, C. B. (1985). "An electron microscopic study of synaptonemal complex formation at zygotene in rye." Chromosoma 92(3): 165-175. Goldman, S. L. et S. Smallets (1979). "Site specific induction of gene conversion: the effects of homozygosity of the ade6 mutant M26 of Schizosaccharomyces pombe on meiotic gene conversion." Mol Gen Genet 173(3): 221-5. Goldstein, P. (1986). "The synaptonemal complexes of Caenorhabditis elegans: the dominant him mutant mnT6 et pachytene karyotype analysis of the X-autosome translocation." Chromosoma 93(3): 256-60. Graubard, M. A. (1932). "Inversion MELANOGASTER." Genetics 17(1): 81-105. in DROSOPHILA

118

119

120 121

122 123

124

125 126

Greenawalt, D. M., X. Cui, et al. (2006). "Strong correlation between meiotic crossovers et haplotype structure in a 2.5-Mb region on the long arm of chromosome 21." Genome Res 16(2): 208-14. Greenwood, T. A., B. K. Rana, et al. (2004). "Human haplotype block sizes are negatively correlated with recombination rates." Genome Res 14(7): 1358-61. Grelon, M., D. Vezon, et al. (2001). "AtSPO11-1 is necessary for efficient meiotic recombination in plants." Embo J 20(3): 589-600. Grey, C., F. Baudat, et al. (2009). "Genome-wide control of the distribution of meiotic recombination." PLoS Biol 7(2): e35. Grishchuk, A. L. et J. Kohli (2003). "Five RecA-like proteins of Schizosaccharomyces pombe are involved in meiotic recombination." Genetics 165(3): 1031-43.

127

128 129 130

154

131

Grishchuk, A. L., R. Kraehenbuehl, et al. (2004). "Genetic et cytological characterization of the RecA-homologous proteins Rad51 et Dmc1 of Schizosaccharomyces pombe." Curr Genet 44(6): 317-28. Guillon, H., F. Baudat, et al. (2005). "Crossover et Noncrossover Pathways in Mouse Meiosis." Mol Cell 20(4): 563-573. Haber, J. (2007). Multiple Mechanisms of Repairing Meganuclease-Induced Double-Strand DNA Breaks in Budding Yeast. Molecular Genetics of Recombination: 285-316. Haber, J. E., P. C. Thorburn, et al. (1984). "Meiotic et mitotic behavior of dicentric chromosomes in Saccharomyces cerevisiae." Genetics 106(2): 185205. Haldane, J. B. S. (1919). "The combination of linkage values, et the calculation of distance between the loci of linked factors." J Genet 8: 299309. Haldane, J. B. S. (1931). "The Cytological Basis of Genetical Interference." Cytologia 3(1): 54-65. Hamer, G., H. Wang, et al. (2008). "Progression of meiotic recombination requires structural maturation of the central element of the synaptonemal complex." J Cell Sci 121(Pt 15): 2445-51. HapMap (2005). "A haplotype map of the human genome." Nature 437(7063): 1299-320. HapMap (2007). "A second generation human haplotype map of over 3.1 million SNPs." Nature 449(7164): 851-61. Hartung, F., S. Suer, et al. (2008). "Topoisomerase 3alpha et RMI1 Suppress Somatic Crossovers et Are Essential for Resolution of Meiotic Recombination Intermediates in Arabidopsis thaliana." PLoS Genet 4(12): e1000285. Hartung, F., S. Suer, et al. (2007). "Two closely related RecQ helicases have antagonistic roles in homologous recombination et DNA repair in Arabidopsis thaliana." Proc Natl Acad Sci U S A 104(47): 18836-41. Hasenkampf, C. A. (1984). "Synaptonemal complex formation in pollen mother cells of Tradescantia." Chromosoma 90(4): 275-284. Hayase, A., M. Takagi, et al. (2004). "A protein complex containing Mei5 et Sae3 promotes the assembly of the meiosis-specific RecA homolog Dmc1." Cell 119(7): 927-40. Hayashi, K., K. Yoshida, et al. (2005). "A histone H3 methyltransferase controls epigenetic events required for meiotic prophase." Nature 438(7066): 374-8.

132 133

134

135

136 137

138 139 140

141

142 143

144

155

145 146

Henderson, S. A. (1963). "Chiasma distribution at diplotene in a locust." Heredity 18(2): 173-190. Henry, J. M., R. Camahort, et al. (2006). "Mnd1/Hop2 facilitates Dmc1dependent interhomolog crossover formation in meiosis of budding yeast." Mol Cell Biol 26(8): 2913-23. Higgins, J. D., S. J. Armstrong, et al. (2004). "The Arabidopsis MutS homolog AtMSH4 functions at an early step in recombination: evidence for two classes of recombination in Arabidopsis." Genes Dev 18(20): 2557-70. Higgins, J. D., E. F. Buckling, et al. (2008). "Expression et functional analysis of AtMUS81 in Arabidopsis meiosis reveals a role in the second pathway of crossing-over." Plant J 54(1): 152-62. Higgins, J. D., E. Sanchez-Moran, et al. (2005). "The Arabidopsis synaptonemal complex protein ZYP1 is required for chromosome synapsis et normal fidelity of crossing over." Genes Dev 19(20): 2488-500. Hillers, K. J. et A. M. Villeneuve (2003). "Chromosome-wide control of meiotic crossing over in C. elegans." Curr Biol 13(18): 1641-7. Hochwagen, A. (2008). "Meiosis." Curr Biol 18(15): R641-R645. Hoffmann, E. R. et R. H. Borts (2004). "Meiotic recombination intermediates et mismatch repair proteins." Cytogenet Genome Res 107(3-4): 232-48. Holliday, R. (1964). "A mechanism for gene conversion in fungi." Genetics Research 5(2): 282-304. Hollingsworth, N. M. et B. Byers (1989). "HOP1: a yeast meiotic pairing gene." Genetics 121(3): 445-62. Holm, P. et S. Rasmussen (1980). "Chromosome pairing, recombination nodules et chiasma formation in diploid Bombyx males." Carlsberg Research Communications 45(6): 483-548. Holm, P. et S. Rasmussen (1983). "Human meiosis VII. Chiasma formation in human spermatocytes." Carlsberg Research Communications 48(4): 415456. Holm, P., S. Rasmussen, et al. (1981). "Chromosome pairing, recombination nodules et chiasma formation in the basidiomycete coprinus cinereus." Carlsberg Research Communications 46(5): 305-346. Housworth, E. A. et F. W. Stahl (2009). "Is There Variation in Crossover Interference Levels Among Chromosomes From Human Males?" Genetics. Hubert, R., M. MacDonald, et al. (1994). "High resolution localization of recombination hot spots using sperm typing." Nat Genet 7(3): 420-4.

147

148

149

150 151 152

153 154 155

156

157

158 159

156

160 161 162 163

Hulten, M. (1974). "Chiasma distribution at diakinesis in the normal human male." Hereditas 76(1): 55-78. Hulten, M. A., R. W. Palmer, et al. (1982). "Chiasma derived genetic maps et recombination fractions: chromosome 1." Ann Hum Genet 46(Pt 2): 167-75. Hunter, N. (2007). Meiotic recombination. Molecular Genetics of Recombination. A. Aguilera et R. Rothstein. Berlin, Springer: 381-442. Hunter, N. et N. Kleckner (2001). "The single-end invasion: an asymmetric intermediate at the double-strand break to double-holliday junction transition of meiotic recombination." Cell 106(1): 59-70. Hurst, D. D., S. Fogel, et al. (1972). "Conversion-associated recombination in yeast." Proc Natl Acad Sci U S A 69(1): 101-5. Ip, S. C., U. Rass, et al. (2008). "Identification of Holliday junction resolvases from humans et yeast." Nature 456(7220): 357-61. Ira, G., A. Malkova, et al. (2003). "Srs2 et Sgs1-Top3 suppress crossovers during double-strand break repair in yeast." Cell 115(4): 401-11. Jackson, N., E. Sanchez-Moran, et al. (2006). "Reduced meiotic crossovers et delayed prophase I progression in AtMLH3-deficient Arabidopsis." Embo J 25(6): 1315-23. Janssens, F. A. (1909). "La thorie de la chiasmatypie. Nouvelle interprtation des cinses de maturation." La Cellule 25: 389-411. Jantsch, V., P. Pasierbek, et al. (2004). "Targeted gene knockout reveals a role in meiotic recombination for ZHP-3, a Zip3-related protein in Caenorhabditis elegans." Mol Cell Biol 24(18): 7998-8006. Jeffreys, A. J., J. K. Holloway, et al. (2004). "Meiotic recombination hot spots et human DNA diversity." Philos Trans R Soc Lond B Biol Sci 359(1441): 141-52. Jeffreys, A. J., L. Kauppi, et al. (2001). "Intensely punctate meiotic recombination in the class II region of the major histocompatibility complex." Nat Genet 29(2): 217-22. Jeffreys, A. J., J. Murray, et al. (1998). "High-resolution mapping of crossovers in human sperm defines a minisatellite-associated recombination hotspot." Mol Cell 2(2): 267-73. Jeffreys, A. J. et R. Neumann (2002). "Reciprocal crossover asymmetry et meiotic drive in a human recombination hot spot." Nat Genet 31(3): 267-71. Jeffreys, A. J. et R. Neumann (2005). "Factors influencing recombination frequency et distribution in a human meiotic crossover hotspot." Hum Mol Genet 14(15): 2277-87.

164 165 166 167

168 169

170

171

172

173 174

157

175 176 177 178 179

Jeffreys, A. J. et R. Neumann (2009). "The rise et fall of a human recombination hot spot." Nat Genet. Jeffreys, A. J., R. Neumann, et al. (2005). "Human recombination hot spots hidden in regions of strong marker association." Nat Genet 37(6): 601-6. Jensen-Seaman, M. I., T. S. Furey, et al. (2004). "Comparative recombination rates in the rat, mouse, et human genomes." Genome Res 14(4): 528-38. John, B. (1976). "Myths et mechanisms of meiosis." Chromosoma 54(4): 295-325. Jones, G. et F. Franklin (2008). Meiosis in Arabidopis thaliana : Recombination, Chromosome Organization et Meiotic Progression. Recombination et Meiosis - Crossing-Over et Disjunction: 279-306. Jones, G. H. (1984). "The control of chiasma distribution." Symp Soc Exp Biol 38: 293-320. Jones, G. H. et J. A. Croft (1986). "Surface spreading of synaptonemal complexes in locusts." Chromosoma 93(6): 489-495. Kauppi, L., M. Jasin, et al. (2007). "Meiotic crossover hotspots contained in haplotype block boundaries of the mouse genome." Proc Natl Acad Sci U S A 104(33): 13396-401. Kauppi, L., A. Sajantila, et al. (2003). "Recombination hotspots rather than population history dominate linkage disequilibrium in the MHC class II region." Hum Mol Genet 12(1): 33-40. Kauppi, L., M. P. Stumpf, et al. (2005). "Localized breakdown in linkage disequilibrium does not always predict sperm crossover hot spots in the human MHC class II region." Genomics 86(1): 13-24. Keeney, S. (2001). "Mechanism et control of meiotic recombination initiation." Curr Top Dev Biol 52: 1-53. Keeney, S. (2008). Spo11 et the Formation of DNA Double-Strand Breaks in Meiosis. Recombination et Meiosis - Crossing-Over et Disjunction: 81123. Keeney, S., C. N. Giroux, et al. (1997). "Meiosis-specific DNA double-strand breaks are catalyzed by Spo11, a member of a widely conserved protein family." Cell 88(3): 375-84. Kemp, B., R. M. Boumil, et al. (2004). "A role for centromere pairing in meiotic chromosome segregation." Genes Dev 18(16): 1946-51. King, J. S. et R. K. Mortimer (1990). "A polymerization model of chiasma interference et corresponding computer simulation." Genetics 126(4): 112738.

180 181 182

183

184

185 186

187

188 189

158

190

Kirkpatrick, D. T., Y. H. Wang, et al. (1999). "Control of meiotic recombination et gene expression in yeast by a simple repetitive DNA sequence that excludes nucleosomes." Mol Cell Biol 19(11): 7661-71. Kleckner, N. (2006). "Chiasma formation: chromatin/axis interplay et the role(s) of the synaptonemal complex." Chromosoma 115(3): 175-94. Kleckner, N., A. Storlazzi, et al. (2003). "Coordinate variation in meiotic pachytene SC length et total crossover/chiasma frequency under conditions of constant DNA length." Trends Genet 19(11): 623-8. Kleckner, N. et B. M. Weiner (1993). "Potential advantages of unstable interactions for pairing of chromosomes in meiotic, somatic, et premeiotic cells." Cold Spring Harb Symp Quant Biol 58: 553-65. Kleckner, N., D. Zickler, et al. (2004). "A mechanical basis for chromosome function." Proc Natl Acad Sci U S A 101(34): 12592-7. Klein, F., P. Mahr, et al. (1999). "A central role for cohesins in sister chromatid cohesion, formation of axial elements, et recombination during yeast meiosis." Cell 98(1): 91-103. Klutstein, M., H. Shaked, et al. (2008). "Functional Conservation of the Yeast et Arabidopsis RAD54-Like Genes." Genetics 178(4): 2389-97. Kneitz, B., P. E. Cohen, et al. (2000). "MutS homolog 4 localization to meiotic chromosomes is required for chromosome pairing during meiosis in male et female mice." Genes Dev 14(9): 1085-97. Kochakpour, N. et P. B. Moens (2008). "Sex-specific crossover patterns in Zebrafish (Danio rerio)." Heredity 100(5): 489-95. Kohli, J. et J. Bahler (1994). "Homologous recombination in fission yeast: absence of crossover interference et synaptonemal complex." Experientia 50(3): 295-306. Kong, A., D. F. Gudbjartsson, et al. (2002). "A high-resolution recombination map of the human genome." Nat Genet 31(3): 241-7. Koszul, R. et N. Kleckner (2009). "Dynamic chromosome movements during meiosis: a way to eliminate unwanted connections?" Trends Cell Biol. Kuroki, Y. et R. Tanaka (1973). "Duration of mitotic cell cycle in the root tip cells of male Rumex acetosa." The Japanese journal of genetics 48(1): 19-26. Lam, S. Y., S. R. Horn, et al. (2005). "Crossover interference on nucleolus organizing region-bearing chromosomes in Arabidopsis." Genetics 170(2): 807-12. Langridge, W. H., T. A. O'Malley, et al. (1970). "Neutral amphiplasty et regulation of the cell cycle in Crepis herbs." Proc Natl Acad Sci U S A 67(4): 1894-900.

191 192

193

194 195

196 197

198 199

200 201 202 203

204

159

205

Laurie, D. A. et M. A. Hulten (1985). "Further studies on chiasma distribution et interference in the human male." Ann Hum Genet 49(Pt 3): 203-14. Laurie, D. A. et G. H. Jones (1981). "Inter-individual variation in chiasma distribution in Chorthippus brunneus (Orthoptera: Acrididae)." Heredity 47(3): 409-416. Laurie, D. A., R. W. Palmer, et al. (1982). "Chiasma derived genetic lengths et recombination fractions: chromosomes 2 et 9." Ann Hum Genet 46(Pt 3): 233-44. Lawrie, N. M., C. Tease, et al. (1995). "Chiasma frequency, distribution et interference maps of mouse autosomes." Chromosoma 104(4): 308-14. Leblon, G. (1972). "Mechanism of gene conversion in Ascobolus immersus I. Existence of a Correlation Between the Origin of Mutants Induced by Different Mutagens et Their Conversion Spectrum." Mol Gen Genet 115(1): 36-48. Lee, J., T. Iwai, et al. (2003). "Temporally et spatially selective loss of Rec8 protein from meiotic chromosomes during mammalian meiosis." J Cell Sci 116(Pt 13): 2781-90. Lenormand, T. (2003). "The evolution recombination." Genetics 163(2): 811-22. of sex dimorphism in

206

207

208 209

210

211 212 213

Lenormand, T. et J. Dutheil (2005). "Recombination difference between sexes: a role for haploid selection." PLoS Biol 3(3): e63. Leu, J. Y., P. R. Chua, et al. (1998). "The meiosis-specific Hop2 protein of S. cerevisiae ensures synapsis between homologous chromosomes." Cell 94(3): 375-86. Lhuissier, F. G., H. H. Offenberg, et al. (2007). "The mismatch repair protein MLH1 marks a subset of strongly interfering crossovers in tomato." Plant Cell 19(3): 862-76. Li, N. et M. Stephens (2003). "Modeling linkage disequilibrium et identifying recombination hotspots using single-nucleotide polymorphism data." Genetics 165(4): 2213-33. Li, W., C. Chen, et al. (2004). "The Arabidopsis AtRAD51 gene is dispensable for vegetative development but required for meiosis." Proc Natl Acad Sci U S A 101(29): 10596-601. Lian, J., Y. Yin, et al. (2008). "Variation in crossover interference levels on individual chromosomes from human males." Hum Mol Genet 17(17): 258394. Lichten, M. et A. S. Goldman (1995). "Meiotic recombination hotspots." Annu Rev Genet 29: 423-44.
160

214

215

216

217

218

219

Lien, S., J. Szyda, et al. (2000). "Evidence for heterogeneity in recombination in the human pseudoautosomal region: high resolution analysis by sperm typing et radiation-hybrid mapping." Am J Hum Genet 66(2): 557-66. Lim, D. S. et P. Hasty (1996). "A mutation in mouse rad51 results in an early embryonic lethal that is suppressed by a mutation in p53." Mol Cell Biol 16(12): 7133-43. Lin, S., R. Cheng, et al. (2001). "Genetic crossover interference in the human genome." Ann Hum Genet 65(Pt 1): 79-93. Lindegren, C. C. et G. Lindegren (1942). "Locally Specific Patterns of Chromatid et Chromosome Interference in Neurospora." Genetics 27(1): 124. Lio, Y. C., A. V. Mazin, et al. (2003). "Complex formation by the human Rad51B et Rad51C DNA repair proteins et their activities in vitro." J Biol Chem 278(4): 2469-78. Liu, Y., M. Tarsounas, et al. (2007). "Role of RAD51C et XRCC3 in genetic recombination et DNA repair." J Biol Chem 282(3): 1973-9. Loidl, J. (1987). "Synaptonemal complex spreading in Allium ursinum: pericentric asynapsis et axial thickenings." J Cell Sci 87 ( Pt 3): 439-48. Loidl, J. et G. Jones (1986). "Synaptonemal complex spreading in Allium. I. Triploid A. spbaerocephalon." Chromosoma 93(5): 420-8. Lopez, E., M. Pradillo, et al. (2008). "Pairing et synapsis in wild type Arabidopsis thaliana." Chromosome Res 16(5): 701-8. Lynn, A., T. Ashley, et al. (2004). "Variation in human meiotic recombination." Annu Rev Genomics Hum Genet 5: 317-49. Lynn, A., K. E. Koehler, et al. (2002). "Covariation of synaptonemal complex length et mammalian meiotic exchange rates." Science 296(5576): 2222-5. Lynn, A., R. Soucek, et al. (2007). "ZMM proteins during meiosis: crossover artists at work." Chromosome Res 15(5): 591-605. Macaisne, N., M. Novatchkova, et al. (2008). "SHOC1, an XPF endonuclease-related protein, is essential for the formation of class I meiotic crossovers." Curr Biol 18(18): 1432-7. Mancera, E., R. Bourgon, et al. (2008). "High-resolution mapping of meiotic crossovers et non-crossovers in yeast." Nature 454(7203): 479-85. Manfrini, N., I. Guerini, et al. (2010). "Processing of meiotic DNA doublestrand breaks requires cyclin-dependent kinase et multiple nucleases." J Biol Chem.

220

221 222

223

224 225 226 227 228 229 230 231

232 233

161

234

Marcou, D., A. Masson, et al. (1979). "Evidence for non-random spatial distribution of meiotic exchanges in Podospora anserina: comparison between linkage groups 1 et 6." Mol Gen Genet 176(1): 67-79. Martini, E., R. L. Diaz, et al. (2006). "Crossover homeostasis in yeast meiosis." Cell 126(2): 285-95. Mather, K. (1936). "The determination of position in crossing-over. I. Drosophila melanogaster." Journal of Genetics 33(2): 207-235. Mather, K. (1938). "Crossing-over." Biological Reviews 13(3): 252-292. Maudlin, I. et E. P. Evans (1980). "Chiasma distribution in mouse oocytes during diakinesis." Chromosoma 80(1): 49-56. May, C., M. Slingsby, et al. (2008). Human Recombination Hotspots: Before et After the HapMap Project. Recombination et Meiosis - Crossing-Over et Disjunction: 195-244. May, C. A., A. C. Shone, et al. (2002). "Crossover clustering et rapid decay of linkage disequilibrium in the Xp/Yp pseudoautosomal gene SHOX." Nat Genet 31(3): 272-5. Mazina, O. M., A. V. Mazin, et al. (2004). "Saccharomyces cerevisiae Mer3 helicase stimulates 3'-5' heteroduplex extension by Rad51; implications for crossover control in meiotic recombination." Cell 117(1): 47-56. McKim, K. S., B. L. Green-Marroquin, et al. (1998). "Meiotic synapsis in the absence of recombination." Science 279(5352): 876-8. McMahill, M. S., C. W. Sham, et al. (2007). "Synthesis-Dependent Strand Annealing in Meiosis." PLoS Biol 5(11): e299. McPeek, M. S. et T. P. Speed (1995). "Modeling interference in genetic recombination." Genetics 139(2): 1031-44. McVean, G. A., S. R. Myers, et al. (2004). "The fine-scale structure of recombination rate variation in the human genome." Science 304(5670): 5814. Mercier, R., S. J. Armstrong, et al. (2003). "The meiotic protein SWI1 is required for axial element formation et recombination initiation in Arabidopsis." Development 130(14): 3309-18. Mercier, R., S. Jolivet, et al. (2005). "Two meiotic crossover classes cohabit in Arabidopsis: one is dependent on MER3,whereas the other one is not." Curr Biol 15(8): 692-701. Meselson, M. S. et C. M. Radding (1975). "A general model for genetic recombination." Proc Natl Acad Sci U S A 72(1): 358-61.

235 236 237 238 239

240

241

242 243 244 245

246

247

248

162

249

Mets, D. G. et B. J. Meyer (2009). "Condensins Regulate Meiotic DNA Break Distribution, thus Crossover Frequency, by Controlling Chromosome Structure." Cell. Meunier, J. et L. Duret (2004). "Recombination drives the evolution of GCcontent in the human genome." Mol Biol Evol 21(6): 984-90. Meuwissen, R. L., H. H. Offenberg, et al. (1992). "A coiled-coil related protein specific for synapsed regions of meiotic prophase chromosomes." Embo J 11(13): 5091-100. Mieczkowski, P. A., M. Dominska, et al. (2006). "Global analysis of the relationship between the binding of the Bas1p transcription factor et meiosis-specific double-strand DNA breaks in Saccharomyces cerevisiae." Mol Cell Biol 26(3): 1014-27. Mikhailova, E. I., D. Phillips, et al. (2006). "Molecular assembly of meiotic proteins Asy1 et Zyp1 et pairing promiscuity in rye (Secale cereale L.) et its synaptic mutant sy10." Genetics 174(3): 1247-58. Mimitou, E. P. et L. S. Symington (2008). "Sae2, Exo1 et Sgs1 collaborate in DNA double-strand break processing." Nature 455(7214): 770-4. Mimitou, E. P. et L. S. Symington (2009). "DNA end resection: many nucleases make light work." DNA Repair (Amst) 8(9): 983-95. Moens, P. B. (2006). "Zebrafish: chiasmata et interference." Genome 49(3): 205-8. Moens, P. B. et M. L. Ashton (1985). "Synaptonemal complexes of normal et mutant yeast chromosomes (Saccharomyces cerevisiae)." Chromosoma 91(2): 113-20. Moens, P. B., D. J. Chen, et al. (1997). "Rad51 immunocytology in rat et mouse spermatocytes et oocytes." Chromosoma 106(4): 207-15. Moens, P. B., N. K. Kolas, et al. (2002). "The time course et chromosomal localization of recombination-related proteins at meiosis in the mouse are compatible with models that can resolve the early DNA-DNA interactions without reciprocal recombination." J Cell Sci 115(Pt 8): 1611-22. Moens, P. B., E. Marcon, et al. (2007). "Initiation et resolution of interhomolog connections: crossover et non-crossover sites along mouse synaptonemal complexes." J Cell Sci 120(Pt 6): 1017-27. Morgan, T. H. (1911). "Random Segregation Versus Coupling in Mendelian Inheritance." Science 34(873): 384. Moses, M. J. (1956). "Chromosomal structures in crayfish spermatocytes." J Biophys Biochem Cytol 2(2): 215-8. Moses, M. J. (1968). "Synaptinemal Complex." Annual Review of Genetics 2(1): 363.
163

250 251

252

253

254 255 256 257

258 259

260

261 262 263

264

Moses, M. J., P. A. Poorman, et al. (1982). "Synaptonemal complex analysis of mouse chromosomal rearrangements. IV. Synapsis et synaptic adjustment in two paracentric inversions." Chromosoma 84(4): 457-74. Muller, H. J. (1916). "The mechanism of crossing-over." Am Nat 50(582): 193-434. Munoz, I. M., K. Hain, et al. (2009). "Coordination of structure-specific nucleases by human SLX4/BTBD12 is required for DNA repair." Mol Cell 35(1): 116-27. Murray, N. E. (1963). "Polarized recombination et fine structure within the me-2 gene of Neurospora crassa." Genetics 48: 1163-83. Myers, S., L. Bottolo, et al. (2005). "A fine-scale map of recombination rates et hotspots across the human genome." Science 310(5746): 321-4. Myers, S., R. Bowden, et al. (2010). "Drive against hotspot motifs in primates implicates the PRDM9 gene in meiotic recombination." Science 327(5967): 876-9. Myers, S., C. Freeman, et al. (2008). "A common sequence motif associated with recombination hot spots et genome instability in humans." Nat Genet 40(9): 1124-1129. Myers, S., C. C. Spencer, et al. (2006). "The distribution et causes of meiotic recombination in the human genome." Biochem Soc Trans 34(Pt 4): 526-30. Myers, S. R. et S. A. McCarroll (2006). "New insights into the biological basis of genomic disorders." Nat Genet 38(12): 1363-4. Nakagawa, T. et R. D. Kolodner (2002). "Saccharomyces cerevisiae Mer3 is a DNA helicase involved in meiotic crossing over." Mol Cell Biol 22(10): 3281-91. Neale, M. J., J. Pan, et al. (2005). "Endonucleolytic processing of covalent protein-linked DNA double-strand breaks." Nature 436(7053): 1053-7. Nicolas, A. (1979). "Variation of gene conversion et intragenic recombination frequencies in the genome of Ascobolus immersus." Mol Gen Genet 176(1): 129-38. Nicolas, A., D. Treco, et al. (1989). "An initiation site for meiotic gene conversion in the yeast Saccharomyces cerevisiae." Nature 338(6210): 35-9. Nimonkar, A. V., R. A. Sica, et al. (2009). "Rad52 promotes second-end DNA capture in double-stranded break repair to form complementstabilized joint molecules." Proc Natl Acad Sci U S A 106(9): 3077-82. Niu, H., L. Wan, et al. (2009). "Regulation of meiotic recombination via Mek1-mediated Rad54 phosphorylation." Mol Cell 36(3): 393-404.

265 266

267 268 269

270

271 272 273

274 275

276 277

278

164

279

Nonomura, K. I., M. Nakano, et al. (2004). "An insertional mutation in the rice PAIR2 gene, the ortholog of Arabidopsis ASY1, results in a defect in homologous chromosome pairing during meiosis." Mol Genet Genomics 271(2): 121-9. Novak, I., H. Wang, et al. (2008). "Cohesin Smc1beta determines meiotic chromatin axis loop organization." J Cell Biol 180(1): 83-90. Offenberg, H. H., J. A. Schalk, et al. (1998). "SCP2: a major protein component of the axial elements of synaptonemal complexes of the rat." Nucleic Acids Res 26(11): 2572-9. Oh, S. D., J. P. Lao, et al. (2007). "BLM ortholog, Sgs1, prevents aberrant crossing-over by suppressing formation of multichromatid joint molecules." Cell 130(2): 259-72. Oh, S. D., J. P. Lao, et al. (2008). "RecQ helicase, Sgs1, et XPF family endonuclease, Mus81-Mms4, resolve aberrant joint molecules during meiotic recombination." Mol Cell 31(3): 324-36. Ohta, K., T. Shibata, et al. (1994). "Changes in chromatin structure at recombination initiation sites during yeast meiosis." Embo J 13(23): 5754-63. Ohta, K., T. C. Wu, et al. (1999). "Competitive inactivation of a doublestrand DNA break site involves parallel suppression of meiosis-induced changes in chromatin configuration." Nucleic Acids Res 27(10): 2175-80. Oliver-Bonet, M., M. Campillo, et al. (2007). "Analysis of replication protein A (RPA) in human spermatogenesis." Mol Hum Reprod 13(12): 837-44. Osakabe, K., K. Abe, et al. (2006). "Isolation et characterization of the RAD54 gene from Arabidopsis thaliana." Plant J 48(6): 827-42. Page, S. L. et R. S. Hawley (2003). "Chromosome choreography: the meiotic ballet." Science 301(5634): 785-9. Page, S. L. et R. S. Hawley (2004). "The genetics et molecular biology of the synaptonemal complex." Annu Rev Cell Dev Biol 20: 525-58. Painter, T. S. (1934). "The Morphology of the X Chromosome in Salivary Glands of Drosophila Melanogaster et a New Type of Chromosome Map for This Element." Genetics 19(5): 448-69. Parvanov, E. D., P. M. Petkov, et al. (2010). "Prdm9 controls activation of mammalian recombination hotspots." Science 327(5967): 835. Pasierbek, P., M. Fodermayr, et al. (2003). "The Caenorhabditis elegans SCC3 homologue is required for meiotic synapsis et for proper chromosome disjunction in mitosis et meiosis." Exp Cell Res 289(2): 245-55. Pasierbek, P., M. Jantsch, et al. (2001). "A Caenorhabditis elegans cohesion protein with functions in meiotic chromosome pairing et disjunction." Genes Dev 15(11): 1349-60.
165

280 281

282

283

284 285

286 287 288 289 290

291 292

293

294

Payne, L. C. (1957). "The theory of genetical recombination: Effect of changing the [frac14][chi]24 intercept length distribution." Heredity 11(1): 129-139. Pecina, A., K. N. Smith, et al. (2002). "Targeted stimulation of meiotic recombination." Cell 111(2): 173-84. Peoples, T. L., E. Dean, et al. (2002). "Close, stable homolog juxtaposition during meiosis in budding yeast is dependent on meiotic recombination, occurs independently of synapsis, et is distinct from DSB-independent pairing contacts." Genes Dev 16(13): 1682-95. Petes, T. D. (2001). "Meiotic recombination hot spots et cold spots." Nat Rev Genet 2(5): 360-9. Petronczki, M., M. F. Siomos, et al. (2003). "Un menage a quatre: the molecular biology of chromosome segregation in meiosis." Cell 112(4): 42340. Petukhova, G. V., R. J. Pezza, et al. (2005). "The Hop2 et Mnd1 proteins act in concert with Rad51 et Dmc1 in meiotic recombination." Nat Struct Mol Biol 12(5): 449-53. Pezza, R. J., O. N. Voloshin, et al. (2007). "Hop2/Mnd1 acts on two critical steps in Dmc1-promoted homologous pairing." Genes Dev 21(14): 1758-66. Pigozzi, M. I. (2001). "Distribution of MLH1 foci on the synaptonemal complexes of chicken oocytes." Cytogenet Cell Genet 95(3-4): 129-33. Pigozzi, M. I. (2008). "Relationship between physical et genetic distances along the zebra finch Z chromosome." Chromosome Res 16(6): 839-49. Pigozzi, M. I. et A. J. Solari (1997). "Extreme axial equalization et wide distribution of recombination nodules in the primitive ZW pair of Rhea americana (Aves, Ratitae)." Chromosome Res 5(6): 421-8. Pigozzi, M. I. et A. J. Solari (1999). "Recombination nodule mapping et chiasma distribution in spermatocytes of the pigeon, Columba livia." Genome 42(2): 308-14. Pittman, D. L., J. Cobb, et al. (1998). "Meiotic prophase arrest with failure of chromosome synapsis in mice deficient for Dmc1, a germline-specific RecA homolog." Mol Cell 1(5): 697-705. Plank, J. L., J. Wu, et al. (2006). "Topoisomerase IIIalpha et Bloom's helicase can resolve a mobile double Holliday junction substrate through convergent branch migration." Proc Natl Acad Sci U S A 103(30): 11118-23. Plug, A. W., A. H. Peters, et al. (1998). "Changes in protein composition of meiotic nodules during mammalian meiosis." J Cell Sci 111 ( Pt 4): 413-23.

295 296

297 298

299

300 301 302 303

304

305

306

307

166

308

Prakash, R., D. Satory, et al. (2009). "Yeast Mph1 helicase dissociates Rad51made D-loops: implications for crossover control in mitotic recombination." Genes Dev 23(1): 67-79. Prieto, I., C. Tease, et al. (2004). "Cohesin component dynamics during meiotic prophase I in mammalian oocytes." Chromosome Res 12(3): 197213. Ptak, S. E., D. A. Hinds, et al. (2005). "Fine-scale recombination patterns differ between chimpanzees et humans." Nat Genet 37(4): 429-34. Ptak, S. E., A. D. Roeder, et al. (2004). "Absence of the TAP2 human recombination hotspot in chimpanzees." PLoS Biol 2(6): e155. Rana, N. A., N. D. Ebenezer, et al. (2004). "Recombination hotspots et block structure of linkage disequilibrium in the human genome exemplified by detailed analysis of PGM1 on 1p31." Hum Mol Genet 13(24): 3089-102. Rasmussen, S. et P. Holm (1978). "Human meiosis II. Chromosome pairing et recombination nodules in human spermatocytes." Carlsberg Research Communications 43(5): 275-327. Rasmussen, S. W. et P. B. Holm (1980). "Mechanics of meiosis." Hereditas 93(2): 187-216. Rattner, J. B., M. R. Goldsmith, et al. (1981). "Chromosome organization during male meiosis in Bombyx mori." Chromosoma 82(3): 341-51. Raynard, S., W. Zhao, et al. (2008). "Functional role of BLAP75 in BLM/Topo IIIalpha -dependent holliday junction processing." J Biol Chem. Revenkova, E., M. Eijpe, et al. (2004). "Cohesin SMC1 beta is required for meiotic chromosome dynamics, sister chromatid cohesion et DNA recombination." Nat Cell Biol 6(6): 555-62. Robine, N., N. Uematsu, et al. (2007). "Genome-wide redistribution of meiotic double-strand breaks in Saccharomyces cerevisiae." Mol Cell Biol 27(5): 1868-80. Rocco, V., B. de Massy, et al. (1992). "The Saccharomyces cerevisiae ARG4 initiator of meiotic gene conversion et its associated double-strand DNA breaks can be inhibited by transcriptional interference." Proc Natl Acad Sci U S A 89(24): 12068-72. Rockmill, B., M. Sym, et al. (1995). "Roles for two RecA homologs in promoting meiotic chromosome synapsis." Genes Dev 9(21): 2684-95. Ross, K. J., P. Fransz, et al. (1996). "A light microscopic atlas of meiosis in Arabidopsis thaliana." Chromosome Res 4(7): 507-16. Saadallah, N. et M. Hulten (1983). "Chiasma distribution, genetic lengths, et recombination fractions: a comparison between chromosomes 15 et 16." Journal of Medical Genetics 20(4): 290-9.
167

309

310 311 312

313

314 315 316 317

318

319

320 321 322

323

Saintenac, C., M. Falque, et al. (2009). "Detailed recombination studies along chromosome 3B provide new insights on crossover distribution in wheat (Triticum aestivum L.)." Genetics 181(2): 393-403. Saito, T. T., T. Tougan, et al. (2004). "Mcp7, a meiosis-specific coiled-coil protein of fission yeast, associates with Meu13 et is required for meiotic recombination." Nucleic Acids Res 32(11): 3325-39. San Filippo, J., P. Sung, et al. (2008). "Mechanism of eukaryotic homologous recombination." Annu Rev Biochem 77: 229-57. Scherthan, H. (2001). "A bouquet makes ends meet." Nat Rev Mol Cell Biol 2(8): 621-7. Scherthan, H., H. Wang, et al. (2007). "Chromosome mobility during meiotic prophase in Saccharomyces cerevisiae." Proc Natl Acad Sci U S A 104(43): 16934-9. Schlecht, H. B., M. Lichten, et al. (2004). "Compartmentalization of the yeast meiotic nucleus revealed by analysis of ectopic recombination." Genetics 168(3): 1189-203. Schmekel, K. et B. Daneholt (1998). "Evidence for close contact between recombination nodules et the central element of the synaptonemal complex." Chromosome Res 6(3): 155-9. Schmuckli-Maurer, J. et W. D. Heyer (2000). "Meiotic recombination in RAD54 mutants of Saccharomyces cerevisiae." Chromosoma 109(1-2): 8693. Schultes, N. P. et J. W. Szostak (1990). "Decreasing gradients of gene conversion on both sides of the initiation site for meiotic recombination at the ARG4 locus in yeast." Genetics 126(4): 813-22. Schwacha, A. et N. Kleckner (1994). "Identification of joint molecules that form frequently between homologs but rarely between sister chromatids during yeast meiosis." Cell 76(1): 51-63. Schwacha, A. et N. Kleckner (1997). "Interhomolog bias during meiotic recombination: meiotic functions promote a highly differentiated interhomolog-only pathway." Cell 90(6): 1123-35. Schweitzer, M. D. (1934). "Coincidence et Interference in Drosophila melanogaster." Proc Natl Acad Sci U S A 20(3): 154-8. Severson, A. F., L. Ling, et al. (2009). "The axial element protein HTP-3 promotes cohesin loading et meiotic axis assembly in C. elegans to implement the meiotic program of chromosome segregation." Genes Dev. Sharan, S. K., A. Pyle, et al. (2004). "BRCA2 deficiency in mice leads to meiotic impairment et infertility." Development 131(1): 131-42.

324

325 326 327

328

329

330

331

332

333

334 335

336

168

337 338 339

Shaw, D. D. et G. R. Knowles (1976). "Comparative chiasma analysis using a computerised optical digitiser." Chromosoma 59(2): 103-27. Sheridan, S. D., X. Yu, et al. (2008). "A comparative analysis of Dmc1 et Rad51 nucleoprotein filaments." Nucleic Acids Res 36(12): 4057-66. Sherman, J. D. et S. M. Stack (1995). "Two-dimensional spreads of synaptonemal complexes from solanaceous plants. VI. High-resolution recombination nodule map for tomato (Lycopersicon esculentum)." Genetics 141(2): 683-708. Shifman, S., J. T. Bell, et al. (2006). "A High-Resolution Single Nucleotide Polymorphism Genetic Map of the Mouse Genome." PLoS Biol 4(12): e395. Shinohara, M., S. L. Gasior, et al. (2000). "Tid1/Rdh54 promotes colocalization of rad51 et dmc1 during meiotic recombination." Proc Natl Acad Sci U S A 97(20): 10814-9. Shinohara, M., S. D. Oh, et al. (2008). "Crossover assurance et crossover interference are distinctly regulated by the ZMM proteins during yeast meiosis." Nat Genet 40(3): 299-309. Shinohara, M., E. Shita-Yamaguchi, et al. (1997). "Characterization of the roles of the Saccharomyces cerevisiae RAD54 gene et a homologue of RAD54, RDH54/TID1, in mitosis et meiosis." Genetics 147(4): 1545-56. Shiroishi, T., T. Sagai, et al. (1993). "Hotspots of meiotic recombination in the mouse major histocompatibility complex." Genetica 88(2-3): 187-96. Siaud, N., E. Dray, et al. (2004). "Brca2 is involved in meiosis in Arabidopsis thaliana as suggested by its interaction with Dmc1." Embo J 23(6): 1392-401. Sigurdsson, S., S. Van Komen, et al. (2001). "Mediator function of the human Rad51B-Rad51C complex in Rad51/RPA-catalyzed DNA strand exchange." Genes Dev 15(24): 3308-18. Smolikov, S., A. Eizinger, et al. (2007). "SYP-3 restricts synaptonemal complex assembly to bridge paired chromosome axes during meiosis in Caenorhabditis elegans." Genetics 176(4): 2015-25. Smolikov, S., K. Schild-Prufert, et al. (2009). "A yeast two-hybrid screen for SYP-3 interactors identifies SYP-4, a component required for synaptonemal complex assembly et chiasma formation in Caenorhabditis elegans meiosis." PLoS Genet 5(10): e1000669. Snowden, T., S. Acharya, et al. (2004). "hMSH4-hMSH5 recognizes Holliday Junctions et forms a meiosis-specific sliding clamp that embraces homologous chromosomes." Mol Cell 15(3): 437-51. Snowden, T., K. S. Shim, et al. (2008). "hMSH4-hMSH5 adenosine nucleotide processing et interactions with homologous recombination machinery." J Biol Chem 283(1): 145-54.
169

340 341

342

343

344 345 346

347

348

349

350

351

Solignac, M., F. Mougel, et al. (2007). "A third generation microsatellitebased linkage map of the honey bee, Apis mellifera, et its comparison with the sequence-based physical map." Genome Biol 8(4): R66. Sonoda, E., M. S. Sasaki, et al. (1998). "Rad51-deficient vertebrate cells accumulate chromosomal breaks prior to cell death." Embo J 17(2): 598-608. Southern, D. I. (1967). "Chiasma distribution in Truxaline grasshoppers." Chromosoma 22(2): 164-191. Spencer, C. C., P. Deloukas, et al. (2006). "The influence of recombination on human genetic diversity." PLoS Genet 2(9): e148. Stack, S. et L. Anderson (1986). "Two-Dimensional Spreads of Synaptonemal Complexes from Solanaceous Plants III. Recombination Nodules et Crossing over in Lycopersicon-Esculentum Tomato." Chromosoma 94(4): 253-258. Stack, S. M. et L. K. Anderson (2002). "Crossing over as assessed by late recombination nodules is related to the pattern of synapsis et the distribution of early recombination nodules in maize." Chromosome Res 10(4): 329-45. Stahl, F. W., H. M. Foss, et al. (2004). "Does crossover interference count in Saccharomyces cerevisiae?" Genetics 168(1): 35-48. Steinmetz, M., Y. Uematsu, et al. (1987). "Hotspots of homologous recombination in mammalian genomes." Trends in Genetics 3: 7-10. Stevens, W. L. (1936). "The analysis of interference." Journal of Genetics 32(1): 51-64. Stone, J. E. et T. D. Petes (2006). "Analysis of the proteins involved in the in vivo repair of base-base mismatches et four-base loops formed during meiotic recombination in the yeast Saccharomyces cerevisiae." Genetics 173(3): 1223-39. Storlazzi, A., S. Gargano, et al. (2010). "Recombination Proteins Mediate Meiotic Spatial Chromosome Organization et Pairing." Cell 141(1): 94-106. Storlazzi, A., S. Tesse, et al. (2003). "Meiotic double-strand breaks at the interface of chromosome movement, chromosome remodeling, et reductional division." Genes Dev 17(21): 2675-87. Stronghill, P. E. et C. A. Hasenkampf (2007). "Analysis of substage associations in prophase I of meiosis in floral buds of wild-type Arabidopsis thaliana (Brassicaceae)." Am. J. Bot. 94(12): 2063-2067. Sturtevant, A. H. (1913). "The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association." J Exp Zool 5(5): 16873. Sturtevant, A. H. (1915). "The behavior of the chromosomes as studied through linkage." Molecular et General Genetics 13(1): 234-287.
170

352 353 354 355

356

357 358 359 360

361 362

363

364

365

366 367

Sun, F., M. Oliver-Bonet, et al. (2004). "Human male recombination maps for individual chromosomes." Am J Hum Genet 74(3): 521-31. Sun, F., M. Oliver-Bonet, et al. (2006). "Variation in MLH1 distribution in recombination maps for individual chromosomes from human males." Hum Mol Genet 15(15): 2376-91. Sun, F., K. Trpkov, et al. (2005). "Variation in meiotic recombination frequencies among human males." Hum Genet 116(3): 172-8. Sun, H., D. Treco, et al. (1989). "Double-strand breaks at an initiation site for meiotic gene conversion." Nature 338(6210): 87-90. Sung, P. (1997). "Function of yeast Rad52 protein as a mediator between replication protein A et the Rad51 recombinase." J Biol Chem 272(45): 28194-7. Sung, P. (1997). "Yeast Rad55 et Rad57 proteins form a heterodimer that functions with replication protein A to promote DNA strand exchange by Rad51 recombinase." Genes Dev 11(9): 1111-21. Surtees, J. A., J. L. Argueso, et al. (2004). "Mismatch repair proteins: key regulators of genetic recombination." Cytogenet Genome Res 107(3-4): 14659. Sutton, W. S. (1902). "On the Morphology of the Chromosome Group in Brachystola magna." Biol Bull 4(1): 24-39. Svendsen, J. M., A. Smogorzewska, et al. (2009). "Mammalian BTBD12/SLX4 assembles a Holliday junction resolvase et is required for DNA repair." Cell 138(1): 63-77. Svetlanov, A., F. Baudat, et al. (2008). "Distinct Functions of MLH3 at Recombination Hot Spots in the Mouse." Genetics 178(4): 1937-45. Sym, M., J. A. Engebrecht, et al. (1993). "ZIP1 is a synaptonemal complex protein required for meiotic chromosome synapsis." Cell 72(3): 365-78. Symington, L. S., A. Brown, et al. (1991). "Genetic analysis of a meiotic recombination hotspot on chromosome III of Saccharomyces cerevisiae." Genetics 128(4): 717-27. Szostak, J. W., T. L. Orr-Weaver, et al. (1983). "The double-strand-break repair model for recombination." Cell 33(1): 25-35. Tarsounas, M., T. Morita, et al. (1999). "RAD51 et DMC1 form mixed complexes associated with mouse meiotic chromosome cores et synaptonemal complexes." J Cell Biol 147(2): 207-20. Tease, C., G. M. Hartshorne, et al. (2002). "Patterns of meiotic recombination in human fetal oocytes." Am J Hum Genet 70(6): 1469-79.

368 369 370

371

372

373 374

375 376 377

378 379

380

171

381

Tease, C. et M. A. Hulten (2004). "Inter-sex variation in synaptonemal complex lengths largely determine the different recombination rates in male et female germ cells." Cytogenet Genome Res 107(3-4): 208-15. Terasawa, M., A. Shinohara, et al. (1995). "Localization of RecA-like recombination proteins on chromosomes of the lily at various meiotic stages." Genes Dev 9(8): 925-34. Tesse, S., A. Storlazzi, et al. (2003). "Localization et roles of Ski8p protein in Sordaria meiosis et delineation of three mechanistically distinct steps of meiotic homolog juxtaposition." Proc Natl Acad Sci U S A 100(22): 1286570. Thijs, H. et C. Heyting (1998). "Polarity of meiotic recombination in the bronze locus of maize." Plant Cell 10(5): 645-8. Thorslund, T., F. Esashi, et al. (2007). "Interactions between human BRCA2 protein et the meiosis-specific recombinase DMC1." Embo J 26(12): 291522. Tiemann-Boege, I., P. Calabrese, et al. (2006). "High-resolution recombination patterns in a region of human chromosome 21 measured by sperm typing." PLoS Genet 2(5): e70. Tishkoff, S. A. et B. C. Verrelli (2003). "Role of evolutionary history on haplotype block structure in the human genome: implications for disease mapping." Curr Opin Genet Dev 13(6): 569-75. Trelles-Sticken, E., J. Loidl, et al. (1999). "Bouquet formation in budding yeast: initiation of recombination is not required for meiotic telomere clustering." J Cell Sci 112 ( Pt 5): 651-8. Tsubouchi, T., A. J. Macqueen, et al. (2008). "Initiation of meiotic chromosome synapsis at centromeres in budding yeast." Genes Dev 22(22): 3217-26. Tsubouchi, T. et G. S. Roeder (2005). "A synaptonemal complex protein promotes homology-independent centromere coupling." Science 308(5723): 870-3. Uematsu, Y., H. Kiefer, et al. (1986). "Molecular characterization of a meiotic recombinational hotspot enhancing homologous equal crossing-over." Embo J 5(9): 2123-9. Vignard, J., T. Siwiec, et al. (2007). "The interplay of RecA-related proteins et the MND1-HOP2 complex during meiosis in Arabidopsis thaliana." PLoS Genet 3(10): 1894-906. Wall, J. D. et J. K. Pritchard (2003). "Haplotype blocks et linkage disequilibrium in the human genome." Nat Rev Genet 4(8): 587-97.

382

383

384 385

386

387

388

389

390

391

392

393

172

394 395 396

Wallace, B. M. et H. Wallace (2003). "Synaptonemal complex karyotype of zebrafish." Heredity 90(2): 136-40. Wang, C. J., P. M. Carlton, et al. (2009). "Interlock Formation et Coiling of Meiotic Chromosome Axes During Synapsis." Genetics. Wang, K., D. Tang, et al. (2009). "MER3 is required for normal meiotic crossover formation, but not for presynaptic alignment in rice." J Cell Sci 122(Pt 12): 2055-63. Wang, T. F., N. Kleckner, et al. (1999). "Functional specificity of MutL homologs in yeast: evidence for three Mlh1-based heterocomplexes with distinct roles during meiosis in recombination et mismatch correction." Proc Natl Acad Sci U S A 96(24): 13914-9. Wang, T. F. et W. M. Kung (2002). "Supercomplex formation between Mlh1-Mlh3 et Sgs1-Top3 heterocomplexes in meiotic yeast cells." Biochem Biophys Res Commun 296(4): 949-53. Webb, A. J., I. L. Berg, et al. (2008). "Sperm cross-over activity in regions of the human genome showing extreme breakdown of marker association." Proc Natl Acad Sci U S A 105(30): 10471-6. Weiner, B. M. et N. Kleckner (1994). "Chromosome pairing via multiple interstitial interactions before et during meiosis in yeast." Cell 77(7): 977-91. Weinstein, A. (1918). "Coincidence of Crossing over in DROSOPHILA MELANOGASTER (AMPELOPHILA)." Genetics 3(2): 135-72. Weinstein, A. (1936). "The Theory of Multiple-Strand Crossing over." Genetics 21(3): 155-99. Wesoly, J., S. Agarwal, et al. (2006). "Differential contributions of mammalian Rad54 paralogs to recombination, DNA damage repair, et meiosis." Mol Cell Biol 26(3): 976-89. White, M. A., M. Dominska, et al. (1993). "Transcription factors are required for the meiotic recombination hotspot at the HIS4 locus in Saccharomyces cerevisiae." Proc Natl Acad Sci U S A 90(14): 6621-5. Winckler, W., S. R. Myers, et al. (2005). "Comparison of fine-scale recombination rates in humans et chimpanzees." Science 308(5718): 107-11. Wojtasz, L., K. Daniel, et al. (2009). "Mouse HORMAD1 et HORMAD2, Two Conserved Meiotic Chromosomal Proteins, Are Depleted from Synapsed Chromosome Axes with the Help of TRIP13 AAA-ATPase." PLoS Genet 5(10): e1000702. Wu, T. C. et M. Lichten (1995). "Factors that affect the location et frequency of meiosis-induced double-strand breaks in Saccharomyces cerevisiae." Genetics 140(1): 55-66.

397

398

399

400 401 402 403

404

405 406

407

173

408 409

Xu, H., M. D. Beasley, et al. (2005). "Absence of mouse REC8 cohesin promotes synapsis of sister chromatids in meiosis." Dev Cell 8(6): 949-61. Xu, L. et N. Kleckner (1995). "Sequence non-specific double-strand breaks et interhomolog interactions prior to double-strand break formation at a meiotic recombination hot spot in yeast." Embo J 14(20): 5115-28. Xu, X., A. P. Hsia, et al. (1995). "Meiotic recombination break points resolve at high rates at the 5' end of a maize coding sequence." Plant Cell 7(12): 2151-61. Yandeau-Nelson, M. D., Q. Zhou, et al. (2005). "MuDR transposase increases the frequency of meiotic crossovers in the vicinity of a Mu insertion in the maize a1 gene." Genetics 169(2): 917-29. Yao, H. et P. S. Schnable (2005). "Cis-effects on meiotic recombination across distinct a1-sh2 intervals in a common Zea genetic background." Genetics 170(4): 1929-44. Yauk, C. L., P. R. Bois, et al. (2003). "High-resolution sperm typing of meiotic recombination in the mouse MHC Ebeta gene." Embo J 22(6): 1389-97. Yildiz, O., S. Majumder, et al. (2002). "Drosophila MUS312 interacts with the nucleotide excision repair endonuclease MEI-9 to generate meiotic crossovers." Mol Cell 10(6): 1503-9. Yip, S. P., J. U. Lovegrove, et al. (1999). "Mapping recombination hotspots in human phosphoglucomutase (PGM1)." Hum Mol Genet 8(9): 1699-706. Yoshida, K., G. Kondoh, et al. (1998). "The mouse RecA-like gene Dmc1 is required for homologous chromosome synapsis during meiosis." Mol Cell 1(5): 707-18. Young, J. A., R. W. Hyppa, et al. (2004). "Conserved et nonconserved proteins for meiotic DNA breakage et repair in yeasts." Genetics 167(2): 593-605. Zetka, M. C., I. Kawasaki, et al. (1999). "Synapsis et chiasma formation in Caenorhabditis elegans require HIM-3, a meiotic chromosome core component that functions in chromosome segregation." Genes Dev 13(17): 2258-70. Zhao, H., T. P. Speed, et al. (1995). "Statistical analysis of crossover interference using the chi-square model." Genetics 139(2): 1045-56. Zickler, D. (2006). "From early homologue recognition to synaptonemal complex formation." Chromosoma 115(3): 158-74. Zickler, D. et N. Kleckner (1998). "The leptotene-zygotene transition of meiosis." Annu Rev Genet 32: 619-97.

410

411

412

413

414

415 416

417

418

419 420 421

174

422 423

Zickler, D. et N. Kleckner (1999). "Meiotic chromosomes: integrating structure et function." Annu Rev Genet 33: 603-754. Zickler, D., P. J. Moreau, et al. (1992). "Correlation between pairing initiation sites, recombination nodules et meiotic recombination in Sordaria macrospora." Genetics 132(1): 135-48. Zickler, D. et J. Sage (1981). "Synaptonemal complexes with modified lateral elements in Sordaria humana: development of et relationship to the recombination nodules." Chromosoma 84(3): 305-18.

424

175

6 Annexe
6.1 La variation des taux de crossing-over le long du chromosome miotique 4 dArabidopsis

thaliana rvle la

prsence de points chauds de recombinaison

176

Letter

Variation in crossing-over rates across chromosome 4 of Arabidopsis thaliana reveals the presence of meiotic recombination hot spots
Jan Drouaud,1 Christine Camilleri,1,2 Pierre-Yves Bourguignon,3 Aurlie Canaguier,1,6 Aurlie Brard,1,2 Daniel Vezon,1 Sandra Giancola,1,2 Dominique Brunel,1,2 Vincent Colot,4 Bernard Prum,3 Hadi Quesneville,5 and Christine Mzard1,7
Station de Gntique et dAmlioration des Plantes, Institut Jean-Pierre Bourgin, Institut National de la Recherche Agronomique (INRA), 78026, Versailles cedex, France; 2INRA/CNG, 91057 Evry cedex, France; 3Laboratoire Statistique et Gnome, UMR 8071 Centre National de la Recherche Scientifique (CNRS)-INRA-Universit Evry Val dEssonne, (UEVE), 91000 Evry, France; 4Unit de Recherche en Gnomique Vgtale (URGV), INRA/CNRS/UEVE, CP5708, 91057 Evry cedex, France; 5Laboratoire de Dynamique du Gnome et Evolution, Institut Jacques Monod, 75251 Paris cedex 05, France Crossover (CO) is a key process for the accurate segregation of homologous chromosomes during the first meiotic division. In most eukaryotes, meiotic recombination is not homogeneous along the chromosomes, suggesting a tight control of the location of recombination events. We genotyped 71 single nucleotide polymorphisms (SNPs) covering the entire chromosome 4 of Arabidopsis thaliana on 702 F2 plants, representing 1404 meioses and allowing the detection of 1171 COs, to study CO localization in a higher plant. The genetic recombination rates varied along the chromosome from 0 cM/Mb near the centromere to 20 cM/Mb on the short arm next to the NOR region, with a chromosome average of 4.6 cM/Mb. Principal component analysis showed that CO rates negatively correlate with the G+C content (P = 3 104), in contrast to that reported in other eukaryotes. COs also significantly correlate with the density of single repeats and the CpG ratio, but not with genes, pseudogenes, transposable elements, or dispersed repeats. Chromosome 4 has, on average, 1.6 COs per meiosis, and these COs are subjected to interference. A detailed analysis of several regions having high CO rates revealed hot spots of meiotic recombination contained in small fragments of a few kilobases. Both the intensity and the density of these hot spots explain the variation of CO rates along the chromosome. [Supplemental material is available online at www.genome.org.]
Meiotic crossovers (COs) and sister chromatid cohesion provide physical links between homologous chromosomes ensuring proper chromosome segregation during the first meiotic division. In most eukaryotes, there is always at least one CO per pair of homologs (obligatory crossover) (Jones 1984, 1987). Cytological, genetic, and molecular studies in many organisms have demonstrated that COs are not evenly distributed along the chromosomes (Jones 1987; Carpenter 1988; Lynn et al. 2002). The tight control of the number and/or localization of COs is crucial. Mutations that reduce CO formation increase chromosome nondisjunction in organisms as diverse as Saccharomyces cerevisiae, Schizosaccharomyces pombe, Caenorhabditis elegans, Drosophila melanogaster (female), Arabidopsis thaliana, and the mouse (for review, see Lynn et al. 2004). In yeast, the distribution of meiotic recombination events (COs and noncrossover gene conversions; NCOs) along chromosomes has been studied in detail by locating DNA double-strand breaks (DSBs), which initiate meiotic recombination (Baudat and Nicolas 1997; Gerton et al. 2000). These studies showed that
Present address: URGV, INRA/CNRS, 2, rue Gaston Crmieux, CP5708, 91057 Evry cedex, France Corresponding author. E-mail mezard@versailles.inra.fr; fax (33) 1 30 83 33 19. Article published online ahead of print. Article and publication date are at http://www.genome.org/cgi/doi/10.1101/gr.4319006.
7 6

DSBs tend to be clustered in chromosomal domains away from telomeres and centromeres (Gerton et al. 2000; Borde et al. 2004). In mammals, COs are also nonrandomly distributed along the chromosomes, with alternate domains having higher or lower levels of recombination (Kong et al. 2002; Nachman 2002). The CO rates tend to be low near the centromeres and increase toward the telomeres. In plants, the CO rates also vary along chromosomes (for review, see Anderson and Stack 2002). In general, centromeric regions have low CO rates compared to telomeric regions. However, in plants, there have been very few highresolution studies in a single chromosome. Many sequence parameters have been linked to the variation of CO rates in eukaryotes. In yeast and mammals, several studies have found a correlation between a high G+C content and a high rate of recombination in large domains (Gerton et al. 2000; Fullerton et al. 2001; Yu et al. 2001; Kong et al. 2002; Petes and Merker 2002; Jensen-Seaman et al. 2004). However, within 23 kb of the recombination initiation site no correlation between the G+C content and the distribution of COs in both yeast and humans was found (for review, see de Massy 2003) and, second, in human, rat, and mouse, when CpG ratio is included in a multiple regression analysis, correlation with the G+C content becomes negative (Kong et al. 2002; Jensen-Seaman et al. 2004). In wheat, barley, and maize, gene-rich regions are more recombinationally active than gene-poor regions (for review, see

16:000000 2006 by Cold Spring Harbor Laboratory Press; ISSN 1088-9051/06; www.genome.org

Genome Research
www.genome.org

177

Drouaud et al.
Schnable et al. 1998). In humans, female CO rates are not correlated with gene density on chromosome 21 (Lynn et al. 2000) whereas male CO rates are correlated, suggesting a different type of control. There are also conflicting results when correlating the density of transposable elements (TEs) and recombination rates (see Wright et al. 2003). Nevertheless, differences in meiotic CO rates between the sexes have been demonstrated in many higher eukaryotes (Lenormand and Dutheil 2005). Therefore, the primary DNA sequence itself cannot explain all of the variation of meiotic recombination. In S. cerevisiae and S. pombe, hot spots have been defined as small DNA fragments of 12 kb, centered around meiotic DSBs that are repaired, using the homologous chromosome, to produce COs or NCOs (Keeney 2001). In mice, humans, and plants, several such regions have been studied in detail and have been found to share common features with hot spots described in yeast. These include high level of COs and NCOs clustered in small segments (12 kb) and a lack of clear consensus sequences (de Massy 2003; Kauppi et al. 2004; Rafalski and Morgante 2004). The distribution of meiotic hot spots along chromosomes is uneven, which suggests a local control of DSB formation. A lot of effort has been made recently to characterize this finescale variation of recombination rates mainly in humans but also in other eukaryotes for various reasons among which is to gain insight into the underlying mechanisms, to assist association studies, or to improve inferences from polymorphism data about selection and population history. However, except for S. cerevisiae, only a few regions have been characterized at the molecular level in other eukaryotes and more genomewide studies are needed to unravel the determinants of hot spot activity. The availability of the Arabidopsis genome sequence (The Arabidopsis Genome Initiative 2000) and the recent development of powerful high-throughput genotyping techniques (Gut 2001; Kwok 2001), allow us to determine precisely the location and rates of COs on one chromosome. Here, we show that CO rates are highly variable on chromosome 4 of Arabidopsis, with some regions having five times more COs than the chromosome average. The CO rates significantly negatively correlate with the G+C content and also significantly correlate with the density of single repeats and with CpG ratio. However, they do not correlate significantly either with genes, pseudogenes, transposable elements, or dispersed repeats. Our data also confirm that COs are subjected to interference on chromosome 4. Finally, we provide evidence of meiotic recombination hot spots and show that both their activity and density contribute to the variation of the CO rates. short arm about 8 Mb long tipped by the nucleolar organizer region (NOR). This region is about 3.64 Mb long and is constituted of almost homogeneous ribosomal DNA repeats (Haberer et al. 1996). The available short arm sequence starts in the last proximal copy of the rDNA repeat (Mayer et al. 1999; The Arabidopsis Information Resource, http://www.arabidopsis.org/). In some accessions, including Columbia (Col) but not Landsberg (Ler), the short arm has a heterochomatic region, called the knob, identified cytologically (Fransz et al. 2000), primarily comprising transposable elements, in which a few genes are insulated (Mayer et al. 1999; Lippman et al. 2004). Moreover, an approximately 1.5-Mb-long region of the short arm, including the knob, is inverted between the two accessions, Col and Ler (Fransz et al. 2000). We genotyped a population of 736 F2 plants resulting from a cross between Col and Ler (see Methods) with 71 SNPs (Supplemental Table 1) chosen from the Monsanto database (Jander et al. 2002) to be evenly spaced on the Arabidopsis chromosome 4. The average interval between two SNPs was 204 kb on the long arm (60 SNPs) and 239 kb on the short arm (11 SNPs).

Variation of CO rates across chromosome 4


After SNP genotyping, we analyzed the variation in CO rates in 702 plants (34 plants had missing data for more than 24 markers and were thus discarded). On average, we genotyped 666 plants (thus representing 1332 meioses because in an F2 plant each chromosome comes from an independent meiosis) per interval. We verified that there was no bias in the segregation of each marker. The cumulated genetic distance of the chromosome was estimated to be 83.9 cM, of which 69 cM corresponded to the long arm (Supplemental Table 2). As the intervals were small, the genetic length of each interval can be simply calculated by dividing the number of recombinant chromosomes by the number of meioses analyzed. Genetic recombination varied greatly along the chromosome, from

Results
Chromosome 4 of A. thaliana is the smallest of its five chromosomes and presents several remarkable features (Fig. 1). It has an acrocentric architecture with a long arm 14.6 Mb long and
Figure 1. Variation of the CO rates on chromosome 4 of A. thaliana. The numbers refer to the intervals given in Supplemental Table 2. The dotted line represents the average CO rate on chromosome 4 (4.6 cM/Mb). A schematic representation of chromosome 4 of A. thaliana is aligned with the diagram. (Black box) NOR (nucleolar organizer region), (gray box) heterochromatic knob, (diamondshaped box) centromere.

Genome Research
www.genome.org

178

Meiotic recombination hot spots in Arabidopsis


0 cM/Mb next to the centromere, to 20.2 cM/Mb next to the NOR (Supplemental Table 2; Fig. 1). The frequencies of COs in different intervals could not be directly compared because of both the variation in interval length and the number of analyzed chromosomes. Therefore, we developed a statistical approach to unambiguously identify intervals that were significantly either colder or hotter than the chromosome average. The approach is based on a simply binomial model of the number of COs in each interval, so that the temperature of an interval is determined by the probability that the number of COs in it exceeds the expected one, under the assumption that the recombination rate is constant along the chromosome. We implemented a statistical program (TETRA) to compute both the average number of COs per nucleotide, and the significance of the observed values from the binomial model (see Methods). TETRA calculated an average of 4.6 108 COs/nucleotide, which is, on average, 1 cM for 217 kb for chromosome 4. Among the 70 intervals tested, TETRA identified 30 intervals with a significant deviation from the average rate of COs; 12 intervals had a significantly lower rate (cold) and 18 had a significantly higher rate (hot) (P > 0.95 and P < 0.05 for the cold and hot intervals, respectively; Supplemental Table 2). The hot intervals were not randomly distributed: four (intervals 6770) were clustered on the short arm next to the NOR and eight (intervals 4356) were clustered in a 3-Mb region on the long arm next to the centromere (Fig. 1). There was almost no genetic recombination in the centromeric and inverted region (intervals 5863) and no clustering of the cold intervals was observed outside the centromeric region. In the middle of the long arm, there were alternate hot and cold intervals, although the temperature of most of these intervals was not significantly different from the chromosome average. In summary, the COs were unevenly distributed along chromosome 4 with alternating hot and mildly cold regions. gene, and TE densities contribute the most to the composition diversity of the intervals (Fig. 2). However, this axis shows that these features do not occur randomly along the chromosome, but follow two opposite gradients: The gene density is low in the pericentromeric and subtelomeric regions and high in the middle of chromosome arms, whereas the opposite is true for pseudogene and TE density. On the second principal component axis, adding 22.6% to the explained variation, the GC content and the CpG ratio appear to be more relevant to CO rates variation. The intervals showing a significantly higher rate of COs tend to cluster regions of low G+C content where the CpG ratio is high. Conversely, the intervals with a low CO rate cluster in regions of high G+C content where the CpG ratio is low (Fig. 2). A regression analysis carried out between the CO rate and the G+C content or CpG ratio confirmed these trends with R2 of 0.18 ( P = 3 10 4 ) for G+C content and an R 2 of 0.20 (P = 1.3 104) for the CpG ratio. The regression was stronger when analyzing only the long arm of the chromosome, with R 2 = 0.36 ( P = 4 10 7 ) for G+C content and R 2 = 0.22 (P = 1.7 104) for the CpG ratio. Of the other regressions tested (gene density, pseudogenes, etc.), only the SSR density had a significant correlation with CO rates (R2 = 0.13; P = 3 103). Therefore, unlike the results obtained in several other eukaryotes, in which a high CO rate tends to correlate with a high G+C content, we suggest that on chromosome 4 of A. thaliana a high

Correlation of CO rates with primary sequence features


We performed a principal component analysis to determine the most relevant genome features that correlated with the observed CO frequencies. The genes, pseudogenes, G+C content, and CpG log ratio, as well as repeated sequences, such as transposable elements (TEs) and single repeats (SSR), were carefully listed from both publicly available data and in-house computed analysis (see Methods). In each interval, we took the G+C content and the CpG ratio (see Methods) and calculated the density of each of the other features. We analyzed the whole chromosome, excluding the intervals 6063 contained in the inverted region. On the first principal component axis, accounting for 46.4% of the variation, we found that gene, pseudo-

Figure 2. Principal component analysis of chromosome 4 of A. thaliana. Numbers refer to the intervals given in Figure 1 and Supplemental Table 2. Hot intervals are indicated in red; cold intervals are indicated in blue.

Genome Research
www.genome.org

179

Drouaud et al.
fall in the 2 + 1 class (that is, we assumed 3 + 0 pairs to be very rare). For the remaining pairs (23), which display four or more CO, we could not attribute CO unambiguously to parental chromosomes, so we discarded them. As expected, one exchange event was the most common occurrence (692 chromatids). Furthermore, we compared the observed distribution of the number of COs to what is expected under a Poisson distribution (Supplemental Table 3), Test of 2 goodness-of-fit shows that the null Poisson hypothesis can be strongly rejected ( 2 = 121.8, P < 5 104). Hence, we can conclude that multiple COs do not occur on chromosome 4 independently one from each other. For each of the 38 plants having two precisely located COs on the same chromatid (Fig. 3, light gray box), we calculated the genetic distance between the two COs. The distance varied from 1.17 to 62.8 cM with a mean distance of 44.1 cM. The mean expected value for randomly distributed double COs was onethird of the chromosome, being 27.9 cM (see Methods). We then classified the 38 plants into four groups: group 1, with events separated by less than 25% of the chromosome (021 cM); group 2, with events separated by more than 25% but less than 50% of the chromosome (2142 cM); group 3, with events separated by more than 50% but less than 75% of the chromosome (4263 cM); and group 4, with events separated by more than 75% of the chromosome (6383.9 cM) (Fig. 4). We compared the observed distribution with the expected distribution if COs were located independently of each other. We found a very strong probability (2 = 27.9, P < 5 103) that double COs were not located independently of each other. The same analysis on only the long arm also showed that the observed distribution and the observed mean distance (36 cM) were very different from the theoretical values (23 cM; data not shown). We then looked at the effect of the centromere on interference. For the 12 chromosomes having one CO on the short arm and the other on the long arm, the mean distance was 59.6 cM, that is, 70% of the genetic length of the chromosome, while the mean distance between two COs occurring on the long arm represents 52% of the genetic length of the long arm. These results confirm that CO location on chromosome 4 is affected by interference and that the centromere is not a barrier to interference.

Figure 3. Number of CO events per chromatid deduced from the genotype of an F2 plant. When an F2 plant displays two COs, either the extremities of both chromosomes are homozygous, and the COs are on one chromatid or both, or the extremities are heterozygous and there is ambiguity between one event on each chromatid or two events on the same chromatid. A similar approach has been used to analyse F2 plants that displays three COs. When an F2 plant displays more than three COs, the recombination history cannot be inferred from the genotype. In the columns COs/chromatid, the numbers in parentheses represent the number of plants in each category. (Light gray box) plants used in the interference analysis. (Dark gray box) plants containing a chromatid with two COs but that could not be used in the interference analysis.

CO rate correlates with a low G+C content. The CpG ratio and SSR density also weakly correlate with CO rates.

Interference on chromosome 4
We obtained 1171 COs for 1404 analyzed meioses. This corresponded to an average of 0.8 events per chromatid and per meiosis, corresponding to 1.6 COs per pair of homologous chromosomes (bivalents) per meiosis. There were, on average, 1.3 events on the long arm and 0.3 events on the short arm. However, if we take into account the 1.5 Mb that are inverted between the two parental lines, and therefore forbidden from forming and/or recovering COs, the ratio of COs per megabase on the short arm was double that of the long arm (0.18 vs. 0.09). For 515 pairs of chromosomes, we were able to determine unequivocally the number of exchanges that each chromatid had undergone (0, 1, or 2 COs) during meiosis (Fig. 3). For 123 pairs of chromosomes harboring two exchanges, we could not unambiguously attribute the recombination events to one or the other chromatid. We reassigned them either to the 1 + 1 or the 2 + 0 class (see Fig. 3) on the basis of the prorata between the sizes of these latter classes in the nonambiguous class with two exchanges. The 41 pairs that exhibit three exchanges that could not be credited to one or the other chromatid were considered to

Figure 4. Distribution of the distances in centiMorgans between double COs. (Histogram in black) observed distribution of double COs in our F2 (see text); (histogram in gray) theoretical distribution of double COs if the position of one CO is independent of the second (Methods).

Genome Research
www.genome.org

180

Meiotic recombination hot spots in Arabidopsis Evidence for the existence of hot spots of recombination
We further investigated several of the 14 intervals having the highest CO rates together with one interval with a slightly above average CO rate and one cold interval. For each interval, we genotyped the corresponding recombinant plants using a set of SNP or indel markers, giving precise locations of the exchange points. We divided the hottest interval (interval 70; Fig. 1) into 15 parts to map the COs at a precision of a few kilobases (Fig. 5A). We found a clearly nonhomogeneous distribution of exchange events. Two very small fragments (3.4 and 3.2 kb) 20 kb apart exhibited a very high rate of COs (>85 cM/Mb), being 15 times higher than the chromosome average (4.6 cM/Mb) and four times higher than the interval average (20.2 cM/Mb). We found that two other fragments in interval 70 had moderately high rates of genetic recombination (40 and 55 cM/Mb, 8 to 10 times the chromosome average). We also analyzed another hot interval (interval 21, Fig. 5B) in the middle of the long arm (Fig. 1). We found one DNA fragment displaying a large increase of genetic recombination in this interval. The recombination rates in the remainder of this interval were mostly lower than the chromosome average. We also observed the same type of spotty CO distribution in the other hot intervals that we investigated (7, 55, 56 and 68, 69; data not shown). We then analyzed interval 57 (Fig. 1), which did not appear to have a significantly high rate of genetic recombination when analyzed by TETRA (6.6 cM/Mb; P = 0.08). We found one DNA fragment of 12 kb displaying a high rate of genetic recombination (40 cM/Mb) whereas the remainder of the interval displayed CO rates below the chromosome average (Fig. 5C). Interval 37 was found to be significantly cold when analyzed by TETRA (2.7 cM/Mb; P > 0.98). We performed the same kind of analysis as for

Figure 5. Fine-scale analysis of the distribution of CO breakpoints in four intervals. (A) interval 70: 44.5kb-3; 16.4kb-2; 6.2kb-5; 19.5kb-4; 17.5kb-3; 9.4kb-4; 12.3kb-7; 8.1kb-0; 8.3kb-2; 3.2kb-4; 13.0kb-5; 6.4kb-1; 3.4kb-5; 3.3kb-0; 11.0kb-7. (B) interval 21: 2.5kb-0; 3.9kb-5; 13.8kb-1; 23.6kb-2; 45.7kb-2; 29.2kb-0; 38.6kb-0. (C) interval 57: 69.5kb-6; 12.2kb-7; 7.1kb-0; 17.4kb-1; 46.5kb-2; 65.0kb-3. (D) interval 37: 35.2kb-2; 32.7kb-0; 38.1kb-1; 8.8kb-2; 12.6kb-1; 28.3kb-1; 20.2kb-1; 49.2kb-2; 38.7kb-1. For each interval the size of each DNA fragment is given and the number of recombinant plants. (Light gray line) G+C content calculated in 1-kb windows. (Black line) CO rates in centiMorgans per megabase. (Small dotted line) interval COs rate average. (Large dotted line) chromosome 4 COs rate average. Above each major peak is the gene organization of the fragment. (Gray box) gene.

Genome Research
www.genome.org

181

Drouaud et al.
the other intervals. We found a dispatch of the 12 CO exchanges in 9 of the 10 fragments studied (Fig. 5D) with a maximum of two events in an 8.8-kb fragment. This small fragment seems to exhibit a slightly higher CO rate than the genome average (Fig. 5D). However, more plants would be needed to confirm this difference. For the four regions analyzed, hot spots did not seem to correlate with G+C content or gene organization (Fig. 5AD). mosomes (i.e., chromosomes 2 and 4) exhibit more interference than the others and suggested that the NOR region itself rather than the size of the chromosome could influence interference. Further analyses are needed to determine whether interference varies along the chromosome, as recently suggested in a study on rice (Esch 2005). Numerous studies have attempted to understand the factors responsible for genetic recombination variations and to identify primary sequence features that may correlate with this variability. In many sexual organisms, such as mammals, birds, yeast, drosophila, and nematodes, positive correlations between the CO rates and G+C content have been observed at the scale of several hundred of kilobases (Hurst et al. 1999; Gerton et al. 2000; Fullerton et al. 2001; Marais et al. 2001; Takano-Shimizu 2001; Yu et al. 2001; Birdsell 2002; Kong et al. 2002; JensenSeaman et al. 2004). Gerton et al. (2000) suggested that regions of high G+C content stimulate recombination. Alternatively, several recent studies have proposed that high levels of recombination may create regions with high G+C content, probably through a biased gene conversion (BGC) toward G+C. In other words, meiotic recombination modifies the base composition through the average density of recombination hot spots (see below; Galtier et al. 2001; Birdsell 2002; Montoya-Burgos et al. 2003; Meunier and Duret 2004). However, in humans, rats, and mice, when CpG ratio is included in a multiple regression model, the correlation with the G+C content becomes negative (Kong et al. 2002; Jensen-Seaman et al. 2004). In contrast, we found that regions of low G+C content and high CpG ratio on chromosome 4 of A. thaliana tend to have higher rates of genetic recombination. Therefore, the BGC hypothesis suggested to explain the correlation found in other eukaryotes may not apply in Arabidopis. However, homologs of genes believed to participate in G+C-biased mismatch repair in other organisms exist in the genome of Arabidopsis (Birdsell 2002). It is also possible that in Arabidopsis BGC cannot affect the nucleotide content due to the high level of inbreeding of the plant that does not favor the formation of heteroduplex DNAs. However, this would explain an absence of correlation but not a negative correlation. In contrast to our results, a study recently reported no correlation between the G+C content and CO rates in Arabidopsis (Marais et al. 2004). We suggest that our observation is due to the higher precision of our recombination map because we studied 702 plants compared to the 101 RILs used in the study of Marais et al. (2004). Therefore, our study in Arabidopsis questions the assumptions made for G+C correlation and so, the problem of causation remains an open query. Data from more species are needed and may reveal a species-specific lineage in the evolution of recombination. A fine-scale analysis of several intervals showed peaks in crossover activity. For example, in the hottest interval (interval 70; Fig. 1), CO breakpoints are found in 12 of the 14 fragments tested, even though there is clustering in two small regions 20 kb apart (Fig. 5A). In other intervals, including one not having a significantly high CO rate, one small DNA fragment accounts for most of the genetic recombination of the interval. In the genome of A. thaliana, this punctuate distribution of CO activity strongly suggests recombination hot spots where recombination events group around an initiation site. In plants, several hot spots of CO activity have been described. The 140-kb a1-sh2 region in maize has peaks of CO activity (three to six times the genome average) in three small intervals (1.73.4 kb) (Yao et al. 2002). Other loci in maize or in rice, such as bronze or waxy, show some properties

Discussion
We obtained a very detailed genetic map of chromosome 4 of A. thaliana by genotyping a series of 71 SNP markers on 702 F2 plants issued from an F1 Col/Ler hybrid. The total size of the genetic map was estimated at 83.9 cM, which is consistent with other maps obtained from crosses of the same accessions: the classical map (76 cM; Meinke et al. 1998), the RFLP map (74.4 cM; Schmidt et al. 1995; Liu et al. 1996), tetrad analysis using the quartet mutation (85 cM; Copenhaver et al. 1998; Lam et al. 2005), and first versions of the RIL genetic map (76 cM; Lister and Dean 1993). We found that, on average, a chromosome 4 bivalent undergoes 1.6 crossovers per meiosis. Copenhaver et al. found an average of 1.5 COs on chromosome 4 in male meiosis in a Col/ Ler cross (Copenhaver et al. 1998; Lam et al. 2005). Meiotic recombination has also been assessed using cytology by recording the numbers and locations of chiasmata on metaphase I bivalents in pollen mother cells of several accessions, including Col and Ler (Sanchez-Moran et al. 2002). Both the genetic and cytological methods gave consistent results, with the mean chiasma frequency being 1.6 for chromosome 4. Therefore, CO frequency on chromosome 4 during meiosis of a Col/Ler F1 hybrid is not greatly different from that in the parents. In most eukaryotes, positive interference (i.e., the probability of COs occurring next to each other is lower than expected) affects the distribution of multiple COs on a single chromosome (see Zickler and Kleckner 1999). However, not all the COs seem to interfere, and recent data suggest two pathways for crossovers in S. cerevisiae, in humans, and in A. thaliana: one pathway being sensitive to interference (class I) and the other insensitive (class II) (Copenhaver et al. 2002; Housworth and Stahl 2003; Higgins et al. 2004; Hollingsworth and Brill 2004; Stahl et al. 2004; Lam et al. 2005; Mercier et al. 2005). We show also that COs are subjected to interference on chromosome 4. Double COs on the same chromatid are significantly further than one-third of the chromosome length apart, contrary to what is expected for randomly distributed COs. In addition, our results suggest that interference is insensitive to centromere as previously proposed by Colombo and Jones (1997) and that the centromere may increase the strength of interference on chromosome 4. However, there is not complete interference, as we observed double COs only a few centiMorgans apart. We could assume, as suggested by previous studies (Copenhaver et al. 2002; Lam et al. 2005) that these close double COs are insensitive to interference and the distant double COs are sensitive to interference. In yeast, the level of interference has been shown to depend on the size of the chromosome with the short chromosomes harboring less interference (Kabback et al. 1999). However, the disparity in size of the chromosomes is less pronounced in Arabidopsis, with the shortest chromosome being more than twothirds of the size of the longest chromosome, while in yeast there is a fourfold difference in size. Moreover, Lam et al. (2005) recently provided evidences that in Arabidopsis NOR-bearing chro-

Genome Research
www.genome.org

182

Meiotic recombination hot spots in Arabidopsis


of hot spots (Dooner and Martinez-Ferez 1997; Okagaki and Weil 1997; Inukai et al. 2000). For the bronze locus, unlike for yeast and mammals, it has been suggested that recombination is initiated uniformly along the gene and not at a preferential site. Therefore, although existence of hot spots seems to be the rule rather than the exception in plants and other higher eukaryotes, there may be some differences. However, all studies on higher eukaryotes looked at no more than one or two intervals that were often selected for a phenotype associated with the recombination event. Therefore, it is difficult to determine whether the observed hot spot patterns in these intervals can be applied to the whole genome. Here, we show that a punctuate distribution of hot spots is a general feature of the chromosome that is not restricted to significantly recombinogenic regions. Our results also strongly suggest that recombination is initiated at preferential sites all along the chromosome. However, both the intensity and the density of the recombination sites influence the variation of recombination, as hot regions contain one or several very hot spots whereas a mildly warm interval would contain only one mild hot spot. Cold regions may contain few spots with a higher rate of recombination than the genome average, but it remains to be demonstrated. A similar result has recently been obtained in the human genome, where strong hot spots have been detected in narrow regions of strong LD and weak hot spots in regions of strong marker association (Jeffreys et al. 2005). We have identified more than 10 small DNA fragments that may behave as hot spots on chromosome 4 of A. thaliana. Further experiments are needed to confirm the strength and the precise location of the initiation site of these hot spots. Their fine characterization and analysis in other genetic backgrounds is needed to determine the factors that govern their activity and distribution.

SNP genotyping
At each of the SNP sites, DNA extracted from the F2 plants was genotyped either by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry, as described by Sauer et al. (2000) or by fluorescence based techniques: the Amplifluor technology (Serological Corporation) or the TaqMan technology (Applied Biosystems). For a number of SNPs, the results obtained with one technique (usually mass spectrometry) were confirmed with one of the two other methods. The techniques used for each SNP are given in Supplemental Table 1.

Statistical analysis of CO rates: TETRA


We define P as the probability of having a CO in a specific position of the chromosome, and assume that P 1. The probability Pi of observing one CO in the ith interval on a chromosome is approximated as PLi, where Li stands for the length of the ith interval. If the number of informative chromosomes for the interval i (i.e., the number of chromosomes for which both SNPs delimiting the interval are available) is written as Vi, then the number, Ni, of COs in the ith interval is distributed according to a binomial B(Vi,PLi) under the null hypothesis that the COs rate is constant along the chromosome. More explicitly, we have for all k: PNi = k =

Vi k

P Lik 1 P LiVik .

According to the observed values, ni, of the number of COs in the different intervals, TETRA computes the average CO rate, P, along the whole chromosome. It then computes the P-value, Ti, of the observed number of COs under the above binomial model, that is: PNi ni =

Methods
F2 recombinant population construction, genomic DNA extraction
The two Arabidopsis accessions, Columbia and Landsberg erecta, were crossed to obtain an F1 hybrid. Self-fertilization from a single F1 was carried out to obtain F2 seeds. Seeds were grown in soil in long-day conditions in the greenhouse. At the rosette stage, the whole material of 736 F2 plants together with plant material from the two parental accessions was collected. DNA was extracted as described (Loudet et al. 2002).

k=ni

Vi

Vi k

P Lik1 P LiVik.

This P-value can be interpreted as the probability that the number of COs in the ith interval exceeds its observed value under the model of homogeneous CO rate along the chromosome.

Statistical analysis of COs interference


We derive the probability distribution function of the distribution of distances between the two COs by assuming that the locations of the two COs are independently uniformly distributed random variables. L is the length of the chromosome, and x and y are locations of the two COs. x and y are uniformly distributed in [0, L]. The distance r = (x y). The distribution of x and y is symmetric under the exchange of x and y; we can condition on x > y. The probability distribution function P(r) is proportional to the length of the segment, in the x, y plane, between the points of coordinates (r, 0) and (L, L r), which is itself proportional to L r. Imposing the normalization of the probability distribution function, we obtain P(r) = 2(1 r/L). The expectation value of r is thus L 0P(r)r dr = L/3. We can deduce the probabilities P1, P2, P3, P4 of r being in each of the four bins [0, L /4], [ L /4, L /2], [ L /2, 3 L /4], [3 L /4, L ]. For instance P1 = 2 1/4 0 (1 x)dx = 7/16. Similarly, we find P2 = 5/16, P3 = 3/16, P4 = 1/16. These probabilities are used in the analysis of Figure 4.

Selection of SNPs
Most of the SNPs were chosen from the Monsanto database (Jander et al. 2002). When convenient SNPs were not found in the database, DNA fragments were amplified at the desired position on the genomic DNA of the two parental accessions and sequenced to identify a SNP suitable for genotyping. A list of the SNPs used in this study is given in Supplemental Table 1. A couple of primers were designed for each SNP to obtain a PCR fragment containing the predicted SNP. A list of the PCR primers used in this study is given in Supplemental Table 1. The PCRs were carried out on the parental accession DNAs using standard conditions: 94C 4 min, (94C 45 sec, 52C 45 sec, 72C 1 min) 35 cycles, 72C, with Eurobio 1 reaction buffer and Taq polymerase. PCR fragments were sequenced (Genome Express) to check the presence and position of the SNPs.

Correlation studies
The A. thaliana genomic sequence and its annotation were downloaded from the TIGR Web site (http://ftp.tigr.org/pub/data/ a_thaliana/ath1/). Gene and pseudogene annotations have been

Genome Research
www.genome.org

183

Drouaud et al.
extracted from TIGR-XML files from release 5 of the genome annotation. Transposable elements have been re-annotated using the RMBLR procedure from the TE annotation pipeline described by Quesneville et al. (2005). The TE reference set used is derived from the A. thaliana RepeatMasker repeat library (March 6, 2004). The same TE family consecutive TE fragments (on both the genome and the reference TE) have been automatically joined if separated by a sequence composed of more than 80% of other TE insertions (in this case we have a nested TE). Otherwise they are joined if a gap of 5000 nucleotides or a region of mismatches 500 nucleotides long separate them. Single repeats (SSR) were found using the Tandem Repeat Finder program (Benson 1999), and repeats by a BLASTN all-byall using BLASTER and GROUPER (Quesneville et al. 2005) without using any simple link clustering coverage constraint. G+C content and CpG were counted with in-house python scripts. CpG ratios were computed by taking the log10 of the C+G dinucleotide frequency divided by the product of G and C frequencies. All data and analysis results were stored in a MySQL database and retrieved by SQL queries. Statistical analyses were carried out using the R software environment (http://cran.r-project.org)
Kucherlapati and G.R. Smith), pp. 529548. American Society for Microbiology, Washington, DC. Colombo, P.C. and Jones, G.H. 1997. Chiasma interference is blind to centromeres. Heredity 79: 214227. Copenhaver, G.P., Browne, W.E., and Preuss, D. 1998. Assaying genome-wide recombination and centromere functions with Arabidopsis tetrads. Proc. Natl. Acad. Sci. 95: 247252. Copenhaver, G.P., Housworth, E.A., and Stahl, F.W. 2002. Crossover interference in Arabidopsis. Genetics 160: 16311639. de Massy, B. 2003. Distribution of meiotic recombination sites. Trends Genet. 19: 514522. Dooner, H.K. and Martinez-Ferez, I.M. 1997. Recombination occurs uniformly within the bronze gene, a meiotic recombination hotspot in the maize genome. Plant Cell 9: 16331646. Esch, E. 2005. Estimation of gametic frequencies from F2 populations using the EM algorithm and its application in the analysis of crossover interference in rice. Theor. Appl. Genet. 111: 100109. Fransz, P.F., Armstrong, S., de Jong, J.H., Parnell, L.D., van Drunen, C., Dean, C., Zabel, P., Bisseling, T., and Jones, G.H. 2000. Integrated cytogenetic map of chromosome arm 4S of A. thaliana: Structural organization of heterochromatic knob and centromere region. Cell 100: 367376. Fullerton, S.M., Carvalho, A.B., and Clark, A.G. 2001. Local rates of recombination are positively correlated with GC content in the human genome. Mol. Biol. Evol. 18: 11391142. Galtier, N., Piganeau, G., Mouchiroud, D., and Duret, L. 2001. GC-content evolution in mammalian genomes: The biased gene conversion hypothesis. Genetics 159: 907911. Gerton, J.L., DeRisi, J., Shroff, R., Lichten, M., Brown, P.O., and Petes, T.D. 2000. Global mapping of meiotic recombination hotspots and coldspots in the yeast Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. 97: 1138311390. Gut, I.G. 2001. Automation in genotyping of single nucleotide polymorphisms. Hum. Mutat. 17: 475492. Haberer, G., Fischer, T.C., and Torres-Ruiz. R.A. 1996. Mapping of the nucleolus organizer region on chromosome 4 in Arabidopsis thaliana. Mol. Gen. Genet. 250: 123128. Higgins, J.D., Armstrong, S.J., Franklin, F.C., and Jones, G.H. 2004. The Arabidopsis MutS homolog AtMSH4 functions at an early step in recombination: Evidence for two classes of recombination in Arabidopsis. Genes & Dev. 18: 25572570. Hollingsworth, N.M. and Brill, S.J. 2004. The Mus81 solution to resolution: Generating meiotic crossovers without Holliday junctions. Genes & Dev. 18: 117125. Housworth, E.A. and Stahl, F.W. 2003. Crossover interference in humans. Am. J. Hum. Genet. 73: 188197. Hurst, L.D., Brunton, C.F., and Smith, N.G. 1999. Small introns tend to occur in GC-rich regions in some but not all vertebrates. Trends Genet. 15: 437439. Inukai, T., Sako, A., Hirano, H.Y., and Sano, Y. 2000. Analysis of intragenic recombination at wx in rice: Correlation between the molecular and genetic maps within the locus. Genome 43: 589596. Jander, G., Norris, S.R., Rounsley, S.D., Bush, D.F., Levin, I.M., and Last, R.L. 2002. Arabidopsis map-based cloning in the post-genome era. Plant Physiol. 129: 440450. Jeffreys, A.J., Neumann, R., Panayi, M., Myers, S., and Donnelly, P. 2005. Human recombination hot spots hidden in regions of strong marker association. Nat. Genet. 37: 601606. Jensen-Seaman, M.I., Furey, T.S., Payseur, B.A., Lu, Y., Roskin, K.M., Chen, C.F., Thomas, M.A., Haussler, D., and Jacob, H.J. 2004. Comparative recombination rates in the rat, mouse, and human genomes. Genome Res. 14: 528538. Jones, G.H. 1984. The control of chiasma distribution. Symp. Soc. Exp. Biol. 38: 293320. Jones, G.H. 1987. Chiasmata. In Meiosis (ed. P.B. Moens), pp. 213244. Academic Press, London. Kaback, D.B., Barber, D., Mahon, J., Lamb, J., and You J. 1999. Chromosome size-dependent control of meiotic reciprocal recombination in Saccharomyces cerevisiae: The role of crossover interference. Genetics 152: 14751486. Kauppi, L., Jeffreys, A.J., and Keeney, S. 2004. Where the crossovers are: Recombination distributions in mammals. Nat. Rev. Genet. 5: 413424. Keeney, S. 2001. Mechanism and control of meiotic recombination initiation. Curr. Top. Dev. Biol. 52: 153. Kong, A., Gudbjartsson, D.F., Sainz, J., Jonsdottir, G.M., Gudjonsson, S.A., Richardsson, B., Sigurdardottir, S., Barnard, J., Hallbeck, B., Masson, G., et al. 2002. A high-resolution recombination map of the human genome. Nat. Genet. 31: 241247. Kwok, P.Y. 2001. Methods for genotyping single nucleotide polymorphisms. Annu. Rev. Genomics Hum. Genet. 2: 235258.

Detection of hot spots


We halved each interval by choosing a convenient SNP (or indel) either from the Monsanto database or by DNA sequencing (see above). We then sequenced the DNA fragment containing the SNP in the recombinant plants and finally distributed the plants according to their genotype within one or the other half interval. This was iteratively repeated until the location of CO breakpoints was obtained within a few kilobases. A list of the SNPs, indels, and corresponding primers is given in Supplemental Table 1. Genomic DNA from plants was amplified by PCR in standard conditions (see above) with an annealing temperature adapted for each set of primers.

Acknowledgments
We thank Mathilde Grelon, Raphal Mercier, Eric Jenczewski, and Valrie Borde for critical reading of the manuscript and Sylvie Jolivet for technical assistance. All the members of the Miose et Recombinaison group provided helpful comments and participated in stimulating discussions. Marc Mzard kindly provided the statistical analysis of interference. This work was supported by grants from the Institut National de la Recherche Agronomique (to C.M.) and the European Union (Epigenome Network of Excellence to V.C.).

References
Anderson, L.K. and Stack, S.M. 2002. Meiotic recombination in plants. Curr. Genomics 3: 507525. The Arabidopsis Genome Initiative 2000. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408: 796815. Baudat, F. and Nicolas, A. 1997. Clustering of meiotic double-strand breaks on yeast chromosome III. Proc. Natl. Acad. Sci. 94: 52135218. Benson, G. 1999. Tandem repeats finder: A program to analyze DNA sequences. Nucleic Acids Res. 27: 573580. Birdsell, J.A. 2002. Integrating genomics, bioinformatics, and classical genetics to study the effects of recombination on genome evolution. Mol. Biol. Evol. 19: 11811197. Borde, V., Lin, W., Novikov, E., Petrini, J.H., Lichten, M., and Nicolas, A. 2004. Association of Mre11p with double-strand break sites during yeast meiosis. Mol. Cell 13: 389401. Carpenter, A.T.C. 1988. Thoughts on recombination nodules, meotic recombination, and chiasmata. In Genetic recombination (eds. E.R.

Genome Research
www.genome.org

184

Meiotic recombination hot spots in Arabidopsis


Lam, S., Horn, S.R., Radford, S.J., Housworth, E.A., Stahl, F.W., and Copenhaver, G.P. 2005. Crossover interference on NOR-bearing chromosomes in Arabidopsis. Genetics 170: 807812. Lenormand, T. and Dutheil, J. 2005. Recombination difference between sexes: A role for haploid selection. PLoS Biol. 3: e63. Lippman, Z., Gendrel, A.V., Black, M., Vaughn, M.W., Dedhia, N., McCombie, W.R., Lavine, K., Mittal, V., May, B., Kasschau, K.D., et al. 2004. Role of transposable elements in heterochromatin and epigenetic control. Nature 430: 471476. Lister, C. and Dean, C. 1993. Recombinant inbred lines for mapping RFLP and phenotypic markers in Arabidopsis thaliana. Plant J. 4: 745750. Liu, Y.G., Mitsukawa, N., Lister, C., Dean, C., and Whittier, R.F. 1996. Isolation and mapping of a new set of 129 RFLP markers in Arabidopsis thaliana using recombinant inbred lines. Plant J. 10: 733736. Loudet, O., Chaillou, S., Camilleri, C., Bouchez, D., and Daniel-Vedele, F. 2002. Bay-0 x Shahdara recombinant inbred line population: A powerful tool for the genetic dissection of complex traits in Arabidopsis. Theor. Appl. Genet. 104: 11731184. Lynn, A., Kashuk, C., Petersen, M.B., Bailey, J.A., Cox, D.R., Antonarakis, S.E., and Chakravarti, A. 2000. Patterns of meiotic recombination on the long arm of human chromosome 21. Genome Res 10: 13191332. Lynn, A., Koehler, K.E., Judis, L., Chan, E.R., Cherry, J.P., Schwartz, S., Seftel, A., Hunt, P.A., and Hassold, T.J. 2002. Covariation of synaptonemal complex length and mammalian meiotic exchange rates. Science 296: 22222225. Lynn, A., Ashley, T., and Hassold, T. 2004. Variation in human meiotic recombination. Annu. Rev. Genomics Hum. Genet. 5: 317349. Marais, G., Mouchiroud, D., and Duret, L. 2001. Does recombination improve selection on codon usage? Lessons from nematode and fly complete genomes. Proc. Natl. Acad. Sci. 98: 56885692. Marais, G., Charlesworth, B., and Wright, S.I. 2004. Recombination and base composition: The case of the highly self-fertilizing plant Arabidopsis thaliana. Genome Biol. 5: R45. Mayer, K. Schuller, C., Wambutt, R., Murphy, G., Volckaert, G., Pohl, T., Dusterhoft, A., Stiekema, W., Entian, K.D., Terryn, N., et al. 1999. Sequence and analysis of chromosome 4 of the plant Arabidopsis thaliana. Nature 402: 769777. Meinke, D.W., Cherry, J.M., Dean, C., Rounsley, S.D., and Koornneef, M. 1998. Arabidopsis thaliana: A model plant for genome analysis. Science 282: 662, 679682. Mercier, R., Jolivet, S., Vezon, D., Huppe, E., Chelysheva, L., Giovanni, M., Nogue, F., Doutriaux, M.P., Horlow, C., Grelon, M., et al. 2005. Two meiotic crossover classes cohabit in Arabidopsis: One is dependent on MER3, whereas the other one is not. Curr. Biol. 15: 692701. Meunier, J. and Duret, L. 2004. Recombination drives the evolution of GC-content in the human genome. Mol. Biol. Evol. 21: 984990. Montoya-Burgos, J.I., Boursot, P., and Galtier, N. 2003. Recombination explains isochores in mammalian genomes. Trends Genet. 19: 128130. Nachman, M.W. 2002. Variation in recombination rate across the genome: Evidence and implications. Curr. Opin. Genet. Dev. 12: 657663. Okagaki, R.J. and Weil, C.F. 1997. Analysis of recombination sites within the maize waxy locus. Genetics 147: 815821. Petes, T.D. and Merker, J.D. 2002. Context dependence of meiotic recombination hotspots in yeast: The relationship between recombination activity of a reporter construct and base composition. Genetics 162: 20492052. Quesneville, H., Bergman, C.M., Andrieu, O., Autard, D., Nouaud, D., Ashburner, M., and Anxolabehere, D. 2005. Combined evidence annotation of transposable elements in genome sequences. PLoS Comput. Biol. 1: e22. Rafalski, A. and Morgante, M. 2004. Corn and humans: Recombination and linkage disequilibrium in two genomes of similar size. Trends Genet. 20: 103111. Sauer, S., Lechner, D., Berlin, K., Lehrach, H., Escary, J.L., Fox, N., Gut, I.G. 2000. A novel procedure for efficient genotyping of single nucleotide polymorphisms. Nucleic Acids Res. 28: E13. Sanchez-Moran, E., Armstrong, S.J., Santos, J.L., Franklin, F.C., and Jones, G.H. 2002. Variation in chiasma frequency among eight accessions of Arabidopsis thaliana. Genetics 162: 14151422. Schmidt, R., West, J., Love, K., Lenehan, Z., Lister, C., Thompson, H., Bouchez, D., and Dean, C. 1995. Physical map and organization of Arabidopsis thaliana chromosome 4. Science 270: 480483. Schnable, P.S., Hsia, A.P., and Nikolau, B.J. 1998. Genetic recombination in plants. Curr. Opin. Plant Biol. 1: 123129. Stahl, F.W., Foss, H.M., Young, L.S., Borts, R.H., Abdullah, M.F., and Copenhaver, G.P. 2004. Does crossover interference count in Saccharomyces cerevisiae? Genetics 168: 3548. Takano-Shimizu, T. 2001. Local changes in GC/AT substitution biases and in crossover frequencies on Drosophila chromosomes. Mol. Biol. Evol. 18: 606619. Wright, S.I., Agrawal, N., and Bureau, T.E. 2003. Effects of recombination rate and gene density on transposable element distributions in Arabidopsis thaliana. Genome Res. 13: 18971903. Yao, H., Zhou, Q., Li, J., Smith, H., Yandeau, M., Nikolau, B.J., and Schnable, P.S. 2002. Molecular characterization of meiotic recombination across the 140-kb multigenic a1-sh2 interval of maize. Proc. Natl. Acad. Sci. 16: 16. Yu, A., Zhao, C.F., Fan, Y., Jang, W.H., Mungall, A.J., Deloukas, P., Olsen, A., Doggett, N.A., Ghebranious, N., Broman, K.W., et al. 2001. Comparison of human genetic and sequence-based physical maps. Nature 409: 951953. Zickler, D. and Kleckner, N. 1999. Meiotic chromosomes: Integrating structure and function. Annu. Rev. Genet. 33: 603754.

Web site references


http://www.arabidopsis.org/; The Arabidopsis information resource. http://ftp.tigr.org/pub/data/a_thaliana/ath1/; Arabidopsis sequence database. http://cran.r-project.org; R software.

Received June 20, 2005; accepted in revised form September 28, 2005.

Genome Research
www.genome.org

185