Paleogenomics in Cereals for Wheat Improvement – PaleoCereal

Paleogenomics, the study of ancestral genome structures (chromosome number, gene content, genome size), allows the identification and characterization of mechanisms (i.e. duplications, translocations and inversions) that have shaped genome species during their evolution. It also permits to define more efficient synteny-based tools for genome mapping based on an accurate knowledge of the relationships between genomes, i.e. cross genome map based cloning. Ancestor genome structure can either be investigated (i) by sequencing fossil DNA of extinct organisms like in mammals or (ii) through modelling, based on large-scale comparative genome analyses (colinearity, conserved synteny) of actual species. In contrast to mammals, paleogenomics has been poorly investigated in plants and particularly in cereals as these species have undergone a large number of whole genome or segmental duplications, diploidization and small-scale rearrangements (translocation, gene conversion) that make high-resolution and accurate comparative studies very challenging. We recently published a paleogenomic analysis in cereals by characterising and combining data from the intra- (duplication characterisation) and inter-specific (synteny characterisation) comparisons based on the rice genome and high resolution EST-derived marker based genetic maps (wheat, maize, sorghum). Analysis of the conservation pattern of shared duplication in the four species led to propose a model for the grass genome evolution from an ancestor with 5 chromosomes (proto-chromosomes) that underwent a whole genome duplication (WGD), 50-70 MYA followed by two interchromosomal translocations and fusions that resulted in a n=12 intermediate ancestor. In this model, rice would have retained the 12 original chromosome number whereas it has been reduced in the other cereal genomes. In wheat, 5 chromosomal fusions resulted in an ancestral wheat genome with n=7 chromosomes. The maize and sorghum genomes have evolved from the intermediate 12 chromosomes ancestor through 2 chromosomal fusions that resulted in an Panicoideae ancestor with n=10 chromosomes. Then, maize and sorghum evolved independently from this n=10 intermediate ancestor. While the sorghum genome structure remained similar to the n=10 chromosome ancestral genome, maize underwent a WGD event, resulting into an intermediate with n=20 chromosomes followed by 17 chromosome fusions to reach its actual n=10 chromosome structure. This evolutionary model is based on the comparison of a single sequenced genome (rice) to high resolution EST-derived marker based genetic maps (wheat, maize, sorghum). The access in 2008 through ongoing collaborations with J. Messing (The Plant Genome Initiative at Rutgers (PGIR), USA) and M Bevan (John Innes Center, UK) to 3 major cereal sequenced genomes (maize 60 284 genes; sorghum 34 008 genes; brachypodium 21 317 genes) offers the opportunity to go beyond the identification of a minimum number of chromosomes through the reconstruction (modelling) of the ancestral gene order onto the 5 proto-chromosomes and the estimation and the ancestral genome size; both features (ancestral gene number/order and genome size) have been the subject of intense speculation in the past ten years (First program output, considered as fundamental research). This ancestor 'physical map' will open then important perspectives in addressing (i) fundamental questions in the field of evolutionary genomics [Which gene functions have been preferentially amplified during evolution' Which gene families have been preferentially impacted neo/sub functionalization' ] as well as (ii) providing valuable genomics tools [Conserved Orthologous Set (COS) markers for fine mapping or candidate genes for major yield (grain/m2, grain/spike, grain size) meta-QTLs conserved at orthologous positions] mainly for the unsequenced Triticeae genomes (wheat, barley) (Second program output, considered as valuable research for wheat varieties improvement). Either the COS markers or candidate genes under ortho-metaQTL will be considered as valuable deliverables (enabling technologies) for ongoing genetic and physical mapping projects in wheat (especially the FP7 program 'Triticeae Genome' leaded by C. Feuillet).