Deep transcriptome sequencing provides new insights into the structural and functional organization of the wheat genome

Abstract

Background: Because of its size, allohexaploid nature, and high repeat content, the bread wheat genome is a good model to study the impact of the genome structure on gene organization, function, and regulation. However, because of the lack of a reference genome sequence, such studies have long been hampered and our knowledge of the wheat gene space is still limited. The access to the reference sequence of the wheat chromosome 3B provided us with an opportunity to study the wheat transcriptome and its relationships to genome and gene structure at a level that has never been reached before.

Results: By combining this sequence with RNA-seq data, we construct a fine transcriptome map of the chromosome 3B. More than 8,800 transcription sites are identified, that are distributed throughout the entire chromosome. Expression level, expression breadth, alternative splicing as well as several structural features of genes, including transcript length, number of exons, and cumulative intron length are investigated. Our analysis reveals a non-monotonic relationship between gene expression and structure and leads to the hypothesis that gene structure is determined by its function, whereas gene expression is subject to energetic cost. Moreover, we observe a recombination-based partitioning at the gene structure and function level.

Conclusions: Our analysis provides new insights into the relationships between gene and genome structure and function. It reveals mechanisms conserved with other plant species as well as superimposed evolutionary forces that shaped the wheat gene space, likely participating in wheat adaptation.

Figure 1

Distribution of the functional regions on the chromosome 3B. (A) Z-score of the feature density of in a 10-Mb sliding window (step 1 Mb) along chromosome 3B. Positive values are in blue, negative values in red. Features include (a) predicted genes, (b) non-expressed predicted genes, (c) expressed predicted genes, (d) NTRs, (e) cis-NATs. The five main regions of the chromosome 3B are depicted at the bottom of the graph: R1 in red; R2a and R2b in blue; C in orange; R3 in green. The borders of these regions are indicated in Mb. (B) Boxplots of the feature density in the five main regions of the chromosome 3B (R1 and R3 in blue; R2a and R2b in yellow; C in red). Features include (a) predicted genes, (b) non-expressed predicted genes, (c) expressed predicted genes, (d) NTRs, (e) cis-NATs.

Figure 2

Distribution of the percentage of transcriptionally active regions expressed in the different number of experimental conditions. Regions were classified according to their expression breadth, that is, the number of conditions in which they were expressed, from 1 to 15. Dark blue: predicted protein-coding genes; red: predicted full genes; green: predicted pseudogenes/gene fragments; purple: syntenic genes; cyan: non-syntenic genes; orange: cis-NATs; light blue: alternative transcripts.

Figure 3

Relationships between gene expression and gene structural and functional features. Expression levels are binned into 30 categories. Each dot is the mean value for genes in the given expression category, and the error bar indicates the standard deviation of the mean.

Figure 4

Distribution and functional partitioning of wheat chromosome 3B. (A) Boxplots of structural and functional gene features in the five main regions of the chromosome 3B (R1 in red; R2a and R2b in blue; C in orange; R3 in green). (a) Gene expression in FPKM; (b) expression breadth in number of conditions; (c) number of alternative transcripts per gene; (d) transcript length; (e) exon number; (f) mean exon length; (g) cumulative intron length; (h) mean intron length. (B) Distribution and segmentation analysis of (a) gene expression in FPKM, (b) expression breadth in number of conditions, (c) number of alternative transcripts per gene, (d) transcript length, (e) exon number, (f) mean exon length, (g) cumulative intron length, (h) mean intron length. Sliding window size: 10 Mb, step: 1 Mb. The five main regions of the chromosome 3B are depicted at the bottom of the graph: R1 in red; R2a and R2b in blue; C in orange; R3 in green.

Figure 5

Distribution of the percentage of genes from three different expression clusters. Z-score of the percentage of expressed genes for a given cluster in a 10-Mb sliding window (step 1 Mb) along chromosome 3B. Positive values are in blue, negative values in red. (a) Constitutively expressed genes; (b) spike- and grain-specific genes; (c) genes preferentially expressed in leaf. The five main regions of the chromosome 3B are depicted at the bottom of the graph: R1 in red; R2a and R2b in blue; C in orange; R3 in green.