Analysis Of Genome-wide Copy Number Variations Among Asian Cultivated Rice (Oryza Sativa L.)

It is generally assumed that different individuals of the same species have very similar genomes.However, there is growing evidence that structural variation in the form of copy number variation (CNV)can lead to variation among the genomes content of individuals. CNVs, DNA polymorphisms such asinsertion/deletions and duplications affecting genome segments larger than1kb, which is thought tocontribute to the extraordinary phenotypic diversity. However, their plasticity of in rice (Oryza sativa L.)has not been investigated on a genomic scale.Here, we used array-based comparative genomic hybridization (aCGH) to create the first genome-widemap of CNVs in rice, revealed a level of structural diversity between the japonica Nipponbare andindica Guangluai4. The analysis of altered segments of DNA estimates that there are more than sixhundred CNV sequences among the two genotypes.We performed further genome-wide analysis of CNVs by comparing20diverse varieties selected fromdifferent groups of Asian cultivated rice relative to Nipponbare. By a detailed selection standard, a totalof2886copy number variable regions conservatively were detected throughout all12chromosomes,covering10.3Mb (~2.7%of the rice genome). Overall2557,276, and53were called as loss, gain andboth loss and gain in these samples, respectively. There were little correlation between CNVsoccurrence and chromosome length. Chromosome11show the greatest enrichment for CNVs.The most common classes of TEs and other repetitive sequences, were found in majority of regions thatwere identified as being deletions relative to the Nipponbare (2175/2557). Although differences in theamounts within TE classes at specific locations varied widely, LTR retrotransposons were thepredominant class followed by DNA TEs.57.8%(1668/2886) of variants were observed only in indica/japonica, the majority of events areobserved in only one sample. Therefore, the subspecies-specific CNVs are enriched for rare variants,and suggesting that these may represent relatively new events that have arisen within breedingpopulations. Moreover, this is consistent with high levels of CNV diversity within subspecies.There were only one variant that were observed in all indica but in none of japonica, and thus there wasno evidence for strong effects of subspecies differentiation on structural variation. It should be note thatCNVs were documented based on comparisons to a reference japonica genome, and that sequencespresent in indica but not in japonica cannot be detected. Thus, subspecies differentiation-associatedCNVs would be expected to be present in most indica, but in few or no japonica.We identified and confirmed marked differences in their CNV frequencies across diverse groups,indicating that some rice CNVs are likely to arise independently in groups and contribute to groupdifferences. Hierarchical clustering analysis of CNV regions displayed some aspects of known ricerelationships and showed that samples mostly tend to group according to their groups. Overall41.6%(1200/2886) of CNV regions affected whole or partial1321rice annotated genes. GOcategories significantly (p<0.001) enriched for specific biological functions, such as cell death, proteinphosphorylation, and defense response. Many of the genes are members of large gene families,suggesting that CNVs may not result in major qualitative variation due to buffering by redundantfunctions encoded elsewhere in the genome, it may significantly contribute to quantitative variation. Inaddition, many of the rice samples were missing inequal numbers of genes relative to Nipponbare, thisindicate that gene content in CNVs have potential contribution to the heterosis during domestication.Moreover, the majority of genes affected by CNVs are conserved in maize and/or Arabidopsis. This isconsistent with the fact that many of the structural variant genes with orthologs in other species arewithin-species duplications that have moved from their ancestral positions.Our results provide a valuable resource beyond microsatellites and single nucleotide polymorphisms(SNPs) to explore the full dimension of genetic variability for future rice genetic resources research. Wediscuss how a catalog of structural variations in rice will accelerate the identification of the genetic basisof traits, beginning with the use of whole genome association and candidate-CNV/gene approaches. Weanticipate that the CNV resources developed in this work will complement existing genome-wide panelsof SNP markers by providing the foundation for future association studies to delimit how structuralvariation contributes to rice phenotypic variation and production increase.