Whole-genome sequencing of Oryza brachyantha reveals mechanisms underlying Oryza genome evolution

Abstract

The wild species of the genus Oryza contain a largely untapped reservoir of agronomically important genes for rice improvement. Here we report the 261-Mb de novo assembled genome sequence of Oryza brachyantha. Low activity of long-terminal repeat retrotransposons and massive internal deletions of ancient long-terminal repeat elements lead to the compact genome of Oryza brachyantha. We model 32,038 protein-coding genes in the Oryza brachyantha genome, of which only 70% are located in collinear positions in comparison with the rice genome. Analysing breakpoints of non-collinear genes suggests that double-strand break repair through non-homologous end joining has an important role in gene movement and erosion of collinearity in the Oryza genomes. Transition of euchromatin to heterochromatin in the rice genome is accompanied by segmental and tandem duplications, further expanded by transposable element insertions. The high-quality reference genome sequence of Oryza brachyantha provides an important resource for functional and evolutionary studies in the genus Oryza.

Figure 1. Alignment of 36 sequence blocks of O. brachyantha to rice chromosomes.

Rice chromosomes are shown on the left, filled with density of repeat elements for every 100 kb window. Sequence blocks of O. brachyantha on each chromosome are represented by black boxes on the right. Syntenic regions, defined as described in Methods, are shaded in grey. Large sequence blocks were anchored to the chromosomes by a cytogenetic approach. Small sequence blocks of pericentromeric regions, in which few collinear genes could be defined, were anchored to the chromosomes by integration with the physical map and confirmation by Southern blot.

Figure 2. Distributions of genomic features in O. brachyantha and O. sativa on chromosome 4.

The two Oryza genomes showed similar distributions of gene and repeat elements: DNA transposons (DNA-TEs) showed a relatively even distribution along the chromosomes, whereas retrotransposons (RTs) were negatively correlated with the distribution of genes. The overall level of RTs was higher in O. sativa than in O. brachyantha, in which the short arms and the proximal regions of the long arms of chromosome 4 showed much obvious contrast. These highly repetitive regions also showed reduced levels of transcription and gene collinearity. The 4',6-diamidino-2-phenylindole-stained pachytene chromosomes for O. brachyantha and O. sativa were put along the genomic tracks of chromosome 4. The black arrows indicate the transition region of euchromatin/heterochromatin, whereas the red arrows indicate the centromere positions.

Figure 3. Dynamic evolution of LTR retrotransposon in O. brachyantha and O. sativa.

(a) Insertion time of LTR retrotransposons in the O. brachyantha and rice genomes. For comparison, LTR retrotransposons of rice chromosomes 3 and 8 were annotated using the same method described for O. brachyantha. The insertion times of LTR retrotransposons were estimated from intact LTR retrotransposons as described in Methods. (b) Divergence time of solo LTR families in O. brachyantha. The divergence times of five families with only solo LTR members were estimated as described in Methods.

Figure 4. Venn diagram showing the distribution of gene families between O. brachyantha, O. sativa and Sorghum bicolor.

Orthologous gene families are defined in Methods. The numbers of gene families and genes clustered in families are indicated for every species. The intersections between species indicate the numbers of shared gene families, whereas the numbers of unique families are shown in species-specific areas.

Figure 5. Comparison of gene families between O. brachyantha and O. sativa.

(a) Size variation of gene families. All shared gene families between O. brachyantha and O. sativa, excluding one-to-one orthologues, were compared to show the variation in gene family sizes. Negative values indicate fewer family members in O. sativa, whereas positive values indicate more members in O. sativa. Os, O. sativa; Ob, O. brachyantha. (b) Copy number variation of conserved functional domains. The gene number belonging to each conserved functional domain was retrieved according to Pfam domain annotation (Methods). The χ²-test, or Fisher’s exact test when the expected frequency was smaller than five, was employed to select significantly different functional domains between O. brachyantha and O. sativa. Multiple comparisons were corrected by the Bonferroni method as implemented in R. Only significant functional domains with a total number of more than 100 members in two Oryza genomes are shown, ranked by P-value (P-value ≤2 × 10⁻⁴ and q-value ≤0.05).

Figure 6. Distribution of sequence rearrangements along the chromosomes of O. sativa.

Segmental duplications, tandem duplications, inversions and non-collinear sequence blocks were detected as described in Methods. (a), Donors and acceptors of segmental duplications on rice chromosomes are connected by red lines. (b), Chromosome ideograms in the inner circle are represented by different colours, with chromosome numbers indicated. The three internal heatmaps represent inversions (c), expansion of tandem gene duplications (d) and expansion of non-collinear sequence blocks (e), in which expansion in rice is indicated by red colour and expansion in O. brachyantha indicated by blue colour. Of the 214 inversions between O. brachyantha and rice genomes, only inversions with more than two collinear genes are shown. Of the 2,460 orthologous tandem gene clusters, 1,378 tandem clusters expanded in the rice genome are shown as red colour and 654 tandem clusters expanded in O. brachyantha are shown as blue colour. Of the non-collinear sequence blocks, only those regions that had expanded more than double in size in rice or O. brachyantha are shown in red or blue, respectively. (f), The densities of genes, RNA retrotransposons and DNA transposons are shown as green, red and orange histograms in the outer circle, respectively. Min–Max: gene (0–25%); RNA retrotransposon (0–80%); DNA transposon (0–80%). Centromeres are indicated by black bars in the outer circle.

Figure 7. Genomic rearrangements and signature of breakpoints.

(a) Sequence analysis of a non-collinear gene locus. The region containing an expansion on chromosome 1 of O. sativa was compared with orthologous regions in O. brachyantha and O. glaberrima. Nearly 17.2 kb of the expansion in O. sativa (acceptor) was formed by an insertion after its split with O. glaberrima. The insertion sequence was highly homologous to a genomic segment on chromosome 9 of O. sativa (donor). Analysis of the breakpoints indicated that the acceptor sequence was a duplicated copy of the donor sequence. (b) Molecular signatures of recently formed non-collinear genes. Red characters, target site duplication; black arrowhead boxes, genes; shaded black arrowhead boxes, pseudogenes; blue boxes, LTR retrotransposons; green boxes, DNA transposons; orange boxes, non-LTR retrotransposons; sequences between square brackets have highly homologous donor sequences; sequences between dashes have no donor sequences. Detailed sequence analyses are shown in Supplementary Figs S17 and S18. MULE, Mutator-like element.